Copyright 2014

by

Janine Micheli-Jazdzewski

ii

Dedication

I would like to dedicate this thesis to Rock, who is not with us anymore, TR, General

Jack D. Ripper, and Page. Thank you for sitting with me while I worked for countless

hours over the years.

iii

Acknowledgements

I would like to express my special appreciation and thanks to my advisor Dr. Deanna

Kroetz, you have been a superb mentor for me. I would like to thank you for encouraging my research and for helping me to grow as a research scientist. Your advice on both research, as well as on my career have been priceless.

I would also like to thank my committee members, Dr. Laura Bull, Dr. Steve Hamilton and Dr. John Witte for guiding my research and expanding my knowledge on statistics, genetics and clinical phenotypes. I also want to thank past and present members of my laboratory for their support and help over the years, especially Dr. Mike Baldwin, Dr.

Sveta Markova, Dr. Ying Mei Liu and Dr. Leslie Chinn. Thanks are also due to my many collaborators that made this research possible including: Dr. Eric Jorgenson, Dr.

David Bangsberg, Dr. Taisei Mushiroda, Dr. Michiaki Kubo, Dr. Yusuke Nakamura, Dr.

Jeffrey Martin, Joel Mefford, Dr. Sarah Shutgarts, Dr. Sulggi Lee and Dr. Sook Wah

Yee. A special thank you to the RIKEN Center for Genomic Medicine that generously performed the genome-wide genotyping for these projects. Thanks to Dr. Steve

Chamow, Dr. Bill Werner, Dr. Montse Carrasco, and Dr. Teresa Chen who started me on the path to becoming a scientist.

Special thanks to my parents, Dr. Robert Micheli (the real Dr. Micheli) and Edie

Micheli and my sister Jill Micheli who have supported and encouraged me to continue my education. Thank you to my husband John Jazdzewski who has put up with me on a day to day basis. I would also like to thank all of my friends (Dana!) who supported me in writing, and encouraged me to strive towards my goal.

iv

Abstract

Since the emergence of the HIV epidemic it has been recognized that complications to HIV infection and variations in drug response and toxicity are influenced by patient genetics. Identification of genetic predictors of HIV infection complications and variation in drug response and toxicity will lead to better treatment options for patients and reduce HIV-related mortality and morbidity. This dissertation contains research that uses candidate gene and genome-wide approaches to identify and characterize novel genetic predictors of nevirapine pharmacokinetics, nucleoside reverse transcriptase inhibitor-induced peripheral neuropathy and HIV-induced peripheral neuropathy. This research demonstrates that nevirapine pharmacokinetic properties are heritable in

European and African patients and characterizes the significant effects of CYP2B6

516G>T, CYP2B6 983T>C and ABCC10 rs2125739 on nevirapine Cmin concentrations in a Ugandan HIV+ population. It also highlights the importance of considering all three polymorphisms for prediction of nevirapine Cmin. This dissertation also explores the genetic predictors of NRTI-SN using whole genome and candidate gene approaches in a Ugandan HIV+ population. A polymorphism in VAMP4, rs188298690, was identified in the whole genome study and bioinformatic analyses found that this marker is in an active regulatory region and also a population specific eQTL locus. The candidate gene analysis found that polymorphisms in SLC28A1 and ABCC4 are predictive of the development of NRTI-SN. Finally, this dissertation describes research to identify genetic predictors of HIV-SN. Several polymorphisms in the FOLH1 region were identified in a whole genome study and bioinformatic analyses support a role for these polymorphisms in determining FOLH1 expression. Analysis of the top FOLH1

v polymorphism in additional samples showed a trend towards significance and a meta- analysis of the discovery and replication cohorts had improved statistical significance.

The research obtained in this dissertation increases the understanding of the role of genetic variation in determining antiviral pharmacokinetics and toxicity and in complications to HIV infection.

vi

TABLE OF CONTENTS

TITLE PAGE ...... i

DEDICATION ...... iii

ACKNOWLEGEMENTS ...... iv

ABSTRACT ...... v

TABLE OF CONTENTS ...... vii

LIST OF TABLES ...... xi

LIST OF FIGURES ...... xiii

Chapter 1: Introduction

1.1. HISTORY OF HIV/AIDS ...... 1

1.2. OVERVIEW OF HIV ...... 6

1.3. AZT DISCOVERY AND APPROVAL ...... 11

1.4. ANTIRETROVIRAL (ARV) PHARMACOLOGY ...... 14

1.4.1. ARV Overview ...... 14

1.4.2. NRTI Pharmacology ...... 18

1.4.3. NNRTI Pharmacology ...... 19

1.4.4. PI Pharmacology ...... 21

1.4.5. Integrase Inhibitor and Entry Inhibitor Pharmacology ...... 22

1.5. PHARMACOGENETICS OF ARV THERAPY ...... 23

1.6. DISSERTATION AIMS ...... 28

1.7. REFERENCES ...... 30

Chapter 2: Measuring the Overall Genetic Component of Nevirapine Pharmacokinetics and the Role of Selected Polymorphisms: Towards Addressing the Missing Heritability in Pharmacogenetic Phenotypes?

2.1. ABSTRACT ...... 41

vii

2.2. INTRODUCTION ...... 43

2.3. MATERIALS AND METHODS ...... 45

2.3.1. Study Design and Subjects ...... 45

2.3.2. Nevirapine Quantification ...... 45

2.3.3. Genotyping ...... 46

2.3.4. Calculation of Pharmacokinetic Parameters ...... 47

2.3.5. Calculation of Relative Genetic Component ...... 47

2.3.6. Statistical Methods ...... 48

2.4. RESULTS ...... 48

2.4.1. Ethnicity does not play a role in nevirapine AUC0-6h variability ...... 48

2.4.2. Age and sex do not play a role in the variability of nevirapine AUC0-6h .... 51

2.4.3. There is a genetic contribution to variation in nevirapine AUC0-6h...... 51

2.4.4. CYP2B6 516G>T may influence nevirapine AUC0-6h ...... 51

2.5. DISCUSSION ...... 56

2.6. CONCLUSIONS ...... 57

2.7. REFERENCES ...... 59

Chapter 3: CYP2B6 and ABCC10 Polymorphisms Influence Nevirapine Exposure in HIV+ Ugandans

3.1. ABSTRACT ...... 62

3.2. INTRODUCTION ...... 64

3.3. MATERIALS AND METHODS ...... 65

3.3.1 Study Design and Patients ...... 65

3.3.2. Nevirapine Quantification ...... 66

3.3.3. Genotyping ...... 67

3.3.4. Statistical Methods ...... 69

3.4. RESULTS ...... 71 viii

3.4.1. Characteristics of Study Participants and Analysis of the Effect of Demographic Characteristics on NVP Cmin ...... 71

3.4.2. Several polymorphisms are associated with NVP Cmin ...... 72

3.4.3. Polymorphisms in CYP2B6, ABCC10 and CYP2C19 have Significant Effects on NVP Cmin ...... 74

3.4.4. CYP2B6 and ABCC10 Composite Genotypes have Significant Effects on NVP Cmin ...... 80

3.5. DISCUSSION ...... 82

3.6. CONCLUSION ...... 85

3.7. REFERENCES ...... 86

Chapter 4: Genetic Predictors of HIV-1 Induced Peripheral Neuropathy in Ugandan HIV-1+ Subjects

4.1. ABSTRACT ...... 90

4.2. INTRODUCTION ...... 92

4.3. MATERIALS AND METHODS ...... 93

4.3.1. Participants ...... 93

4.3.2. Genotyping ...... 94

4.3.3. Phenotype ...... 98

4.3.4. Statistical Analyses ...... 99

4.3.5. Bioinformatic Analyses ...... 99

4.4. RESULTS ...... 100

4.4.1. Demographic Data ...... 100

4.4.2. Loci in Chromosome 11 are associated with HIV-induced peripheral neuropathy ...... 101

4.4.3. Polymorphisms on chromosome 11 may have an effect on FOLH1 regulation and expression ...... 107

4.4.4. Replication Results ...... 113

ix

4.5. DISCUSSION ...... 115

4.6. CONCLUSION ...... 117

4.7. REFERENCES ...... 118

Chapter 5: Genetic Predictors of NRTI Induced Peripheral Neuropathy in Ugandan HIV-1+ Subjects

5.1. ABSTRACT ...... 123

5.2. INTRODUCTION ...... 125

5.3. MATERIALS AND METHODS: ...... 126

5.3.1. Participants ...... 126

5.3.2. Genotyping ...... 126

5.3.3. Phenotype ...... 127

5.3.4. Statistical Analyses ...... 128

5.3.5. Candidate Gene Analysis ...... 129

5.3.6. Bioinformatic Analyses ...... 130

5.4. RESULTS ...... 131

5.4.1. Demographic Data ...... 131

5.4.2. A loci on chromosome 1 is associated with NRTI-SN ...... 132

5.4.3. The rs188298690 polymorphism is in an active regulatory region and is located within a VAMP4 eQTL ...... 137

5.4.4. Replication Results ...... 140

5.4.5. Candidate gene study reveals an association of ABCC4 and NRTI-SN 142

5.5. DISCUSSION ...... 145

5.6. CONCLUSION ...... 148

5.7. REFERENCES ...... 149

Chapter 6: Conclusions ...... 155

6.1. REFERENCES ...... 160

x

LIST OF TABLES

Table 1.1. HIV proteins organized by their size function and location ...... 7

Table 1.2. Antiretroviral (ARV) drugs approved by the FDA for the treatment of HIV

infection...... 17

Table 1.3. Selected polymorphisms in drug metabolizing enzymes and drug

transporters associated with ARV response and toxicity...... 26

Table 2.1. Patient demographics and relative genetic contribution (rGC) to nevirapine

AUC0-6h ...... 49

Table 2.2. The effect of ethnicity and genotype on nevirapine exposure ...... 55

Table 3.1. Patient Demographics and Effect on NVP Cmin ...... 72

Table 3.2. Linkage Disequilibrium Filtered Top Variants Associated with NVP Cmin ..... 73

Table 3.3. Relationship Between Genotypic Variants and NVP Cmin ...... 74

Table 3.4. Multivariate analysis of the association of NVP Cmin with Genotypes and

Demographic Covariates ...... 80

Table 4.1. Quality Control of Genotype and Subject Data ...... 95

Table 4.2. Quality Control of Genotype Data ...... 96

Table 4.3. SNPs selected for replication...... 98

Table 4.4. Patient Demographic Data in the Discovery and Replication Cohorts ...... 100

Table 4.5. Top Variants Associated with HIV Induced Peripheral Neuropathy ...... 104

Table 4.6. Results from replication and meta-analyses ...... 114

Table 5.1. SNPs Selected for Replication ...... 127

Table 5.2. Candidate genes with NRTI transport or functional evidence ...... 130

Table 5.3. Patient Demographic Data in the Discovery and Replication Cohorts ...... 131

xi

Table 5.4. Top Variants Associated with NRTI-Induced Peripheral Neuropathy ...... 135

Table 5.5. Results from replication and meta-analyses ...... 141

Table 5.6. Top Candidate Gene Variants Associated with NRTI-Induced Peripheral

Neuropathy ...... 142

xii

LIST OF FIGURES

Figure 1.1. Epidemiologic Aspects of the Current Outbreak of Kaposi's Sarcoma and

Opportunistic Infections...... 2

Figure 1.2. Control Study of Kaposi’s Sarcoma and Peumocystis carinii Pneumonia in

Homosexual Men ...... 3

Figure 1.3. Scanning electron micrograph of an HIV infected H9 T-cell ...... 4

Figure 1.4. Abbott HTLV-III EIA ...... 5

Figure 1.5. Structure of the HIV-1 HXB2 genome...... 7

Figure 1.6. HIV life cycle with descriptions of each step...... 8

Figure 1.7. Structure of AZT...... 11

Figure 1.8. Vials of AZT, first marketed as Retrovir...... 14

Figure 1.9. Numbers of persons infected with HIV, AIDS diagnoses and deaths from

1981-2008...... 15

Figure 2.1. Interindividual variation in nevirapine plasma levels...... 50

Figure 2.2. Nevirapine plasma concentrations following a single oral dose...... 53

Figure 2.3. Nevirapine plasma concentrations following a single oral dose...... 54

Figure 3.1. PharmGKB.org NVP pharmacokinetic pathway...... 68

Figure 3.2. Flow chart of statistical analyses...... 70

Figure 3.3. Nevirapine Cmin as a function of CYP2B6 983T>C genotype...... 75

Figure 3.4. Nevirapine Cmin as a function of ABCC10 rs2125739...... 76

Figure 3.5. Nevirapine Cmin as a function of CYP2B6 516G>T genotype...... 77

Figure 3.6. Nevirapine Cmin as a function of CYP2C19 rs491623 genotype...... 78

Figure 3.7. Nevirapine Cmin as a function of CYP2C19 rs4388808 genotype...... 79

xiii

Figure 3.8. Nevirapine Cmin as a function of CYP2B6 composite genotype...... 81

Figure 3.9. Nevirapine Cmin as a function of CYP2B6/ABCC10 composite genotype. ... 82

Figure 4.1. Principal component analysis of study samples compared to world HAPMAP

populations...... 97

Figure 4.2. Manhattan plot showing the distribution along the human autosomes of

-log10 (P values) obtained for SNP association with HIV-SN cases versus

control subjects...... 102

Figure 4.3. Observed versus expected p-values (-log base 10 scale) for SNP

association with HIV-SN cases versus control subjects...... 103

Figure 4.4. UCSC genome browser image showing position of SNPs associated with

HIV-SN susceptibility in the vicinity of the FOLH1 gene...... 109

Figure 4.5. Loci proximal to FOLH1 associated with HIV-SN...... 110

Figure 4.6. Plot of the association of FOLH1 SNP rs7925419 with HIV-SN case and

control status in HIV+ Ugandan subjects...... 111

Figure 4.7. The effect of rs7925419 genotype on RNA expression of FOLH1 in brain

spinal cord cervical c1 tissue...... 112

Figure 5.1. NRTI-SN case/control definition method...... 128

Figure 5.2. Manhattan plot showing the distribution along the human autosomes of

-log10 (P values) obtained for SNP association with NRTI-SN cases versus

control subjects...... 133

Figure 5.3. Observed versus expected p-values for SNP association with NRTI-SN

cases versus control subjects...... 134

Figure 5.4. UCSC genome browser image showing position of SNPs associated with

xiv

NRTI-SN susceptibility in the vicinity of the VAMP4 gene...... 136

Figure 5.5. SNPs that are in LD with rs188298690 are in regions that have active

regulatory elements...... 137

Figure 5.6. Genevar VAMP4 eQTL LCL data for three African populations...... 138

Figure 5.7. The association of intergenic SNP rs188298690 with NRTI-SN case and

control status in HIV+ Ugandan subjects...... 139

Figure 5.8. UCSC genome browser image showing position of the top candidate SNP

associated with NRTI-SN susceptibility in the vicinity of ABCC4...... 143

Figure 5.9. The effect of ABCC4 SNP rs7317112 genotype on RNA expression in nerve

tissue...... 144

xv

Chapter 1

1. Introduction1.

1.1 History of HIV/AIDS: Initial Epidemic and Virus Discovery

From October 1980 to May 1981, Dr. Michael Gottlieb and colleagues observed five unusual cases of Pneumocystis carinii pneumonia (PCP), an infection most commonly seen in immunosuppressed patients, in previously healthy young homosexual men.

This prompted the Centers for Disease Control (CDC) to issue a report on these cases in their Morbidity and Mortality Weekly Report (MMWR) on June 5, 19811. Two of the five patients had already died at the time the report was issued. Soon after this report was issued the CDC began to receive accounts from other locations of other cases of opportunistic infections and Kaposi’s Sarcoma2. By July 1981 the CDC had identified

108 cases of Kaposi’s Sarcoma and PCP with an onset between January 1976 and July

1981. The Case Fatality Rate (CFR) was 40%, on par with yellow fever2,3. The first published description of the patients suffering from this new immunodeficiency disease appeared in the Lancet in September 19814. Hymes et.al. described eight previously healthy patients suffering from Kaposi’s Sarcoma. What made these cases unique was that the patients were relatively young (median age 34 years old), homosexual and suffered from a type of aggressive Kaposi’s Sarcoma more commonly seen in Africa rather than Europe and North America. Five out of eight of those patients had died at the time of publication. In December 1981, several reports were published in The New

England Journal of Medicine detailing outbreaks of Kaposi’s Sarcoma and PCP in

1

California and New York5-8. These reports also detailed abnormalities in immune function in patients with Kaposi’s Sarcoma or other opportunistic infections.

By January 1982 the newly created CDC Task Force on Kaposi’s Sarcoma and

Opportunistic Infections had published a special report detailing the state of the outbreak of this unknown immune disease9. They identified 216 cases of Kaposi’s

Sarcoma or other opportunistic infections with a mortality rate of 40%. The report highlighted the rapid increase in cases and high mortality rate of those with the disease

(Figure 1)9. This report also discussed that the outbreak seemed to be confined geographically around New York, Los Angeles and San Francisco. The outbreak also was most common in young (median age 35 years), Caucasian (70%) homosexual men

(92%)9.

Figure 1.1 Taken from the CDC’s Special Report - Epidemiologic Aspects of the Current Outbreak of Kaposi's Sarcoma and Opportunistic Infections showing the incidence of cases by month of onset9. Fill within the bars represent cases of Kaposi’s Sarcoma (white), Pneumocystis carinii Pneumonia (diagonal hashes), both (grey) or other infections (black).

2

In September of 1982, the CDC issued an update on what was now being called acquired immune deficiency syndrome (AIDS), highlighting that in addition to homosexual men, cases were being identified in intravenous drug users, Haitians and two patients with hemophilia A10. The CDC then conducted a large scale case-control epidemiologic study to identify risk factors for acquiring AIDS and to characterize laboratory abnormalities in persons with AIDS11,12. These studies suggested that the infectious agent for AIDS would be found in blood or sexual fluid secretions and also characterized differences in immune function in patients compared to controls, specifically differences in T-cell populations and mitogen response (Figure 1.2)12.

2500 *

2000

*

1500

* Normal Values Combined Controls 1000 Number/mm3 Cases

500

0 Total B T T-Helper T-Supressor Lymphocytes

Figure 1.2. Taken from the National Case-Control Study of Kaposi’s Sarcoma and Peumocystis carinii Pneumonia in Homosexual Men: Part 2, Laboratory Results. The graph shows the absolute number of lymphocytes per mm3. Bars denote mean ± SE. *Combined controls are significantly different from cases.

3

Once the CDC and the medical community realized an epidemic was emerging, efforts began to identify the infectious agent responsible for AIDS. In 1983 and 1984 three independent groups isolated what they believed to be the virus responsible for causing AIDS13-15. While there is some controversy surrounding what group isolated the virus first, the viruses they identified all belonged to a family of retroviruses known as Human T-cell Leukemia Virus (HTLV) and were designated by the discoverers

HTLV-III, Lymphadenopathy Associated Virus (LAV) and AIDS associated retroviruses .

Figure 1.3 is an image of a T-cell infected with HIV, showing viral particles (in yellow) budding off the host cell16.

Figure 1.3. Scanning electron micrograph of an HIV infected H9 T-cell. Viral particles can be seen in yellow budding off the infected host cell16.

4

The discovery of the virus that causes AIDS, which would be named Human

Immunodeficiency Virus (HIV) in 1986, allowed researchers to begin focusing on developing diagnostic tests and treatments for HIV. Developing a diagnostic test was critical to protect the blood supply and to identify new cases. Researchers developing diagnostics for the detection of HIV focused on finding antibodies reactive to HIV antigens with favorable sensitivity and specificity profiles17,18. The first diagnostic test approved by the FDA was an ELISA developed by Abbott in March 1985; the CDC issued a recommendation for universal testing of the blood supply shortly thereafter to prevent transmission from tainted blood transfusions (Figure 1.4)19,20. Efforts to create a vaccine against HIV were also undertaken, however, none to date have been shown to be effective against HIV infection21.

Figure 1.4. Abbott HTLV-III EIA. This is the first diagnostic test approved by the FDA for the detection of HIV. From the National Museum of American History, Kenneth E. Behring Center. ID# 2007.0060.001 Courtesy of the National Museum of American History, Smithsonian Institution, Division of Medicine and Science. http://americanhistory.si.edu/collections/search/object/nmah_1322289 5

1.2 Overview of HIV: Origins, Life Cycle and Disease Natural History

Since the first reports of AIDS in 1981, over 25 million people worldwide have died from HIV infections22. Currently, 35.3 million people are living with HIV infections, mostly in resource poor settings such as sub-Saharan Africa22.

Genetic studies have determined that the most common strain worldwide of HIV, group M, originated from a zoonotic transmission from chimpanzee Pan troglodytes troglodytes to humans sometime around 1920 in West-Central Africa23. The factors that allowed HIV to become a pandemic are varied and somewhat disputed, with leading theories including the prevalence of unsterile injections in colonial Africa in the early 20th century, increased urbanization in early 20th century Africa and increased exposure to primates due to increased hunting of bushmeat23-25. HIV is a genetically diverse virus; to date four groups of HIV-1 and eight groups of HIV-2 have been discovered which arose independently in separate zoonotic transmissions with different prevalence rates worldwide23. Within each of those groups there are multiple subtypes and recombinant strains consisting of multiple subtypes23. Globally in HIV-1, subtype C of group M is the most prevalent, whereas in the United States subtype B of group M is the most prevalent23. It is important to note that amongst the different strains of HIV, there are varying rates of disease progression and resistance to drug therapy26,27.

HIV belongs to the genus Lentivirus which is part of the family Retroviridae28.

Lentiviruses are single stranded, positive sense, enveloped RNA viruses. The HIV-1

HXB2 genome, which is considered the reference genome, consists of 9719 base pairs

(bp) and encodes 11 proteins (Figure 1.5 and Table 1)29.

6

Figure 1.5. Structure of the HIV-1 HXB2 genome29.

Table 1.1 HIV proteins organized by their size function and location. Proteins that are drug targets are in bold. Adapted from Leitner et al.29 Name Size Function Localization Gag membrane anchoring; env interaction; MA p17 virion nuclear transport of viral core CA p24 core capsid virion NC p7 nucleocapsid, binds RNA virion p6 binds Vpr virion Pol Protease (PR) p15 Gag/Pol cleavage and maturation virion Reverse reverse transcription, RNAse H Transcriptase p66,p51 virion activity (RT) RNase H p15 virion Integrase (IN) p31 DNA provirus integration virion external viral glycoproteins bind to plasma membrane, virion Env gp120/gp41 CD4 and secondary receptors envelope Tat p16/p14 viral transcriptional transactivator primarily in nucleolus/nucleus primarily in nucleolus/nucleus RNA transport, stability and utilization Rev p19 shuttling between nucleolus factor (phosphoprotein) and cytoplasm promotes virion maturation and Vif p23 cytoplasm, virion infectivity promotes nuclear localization of Vpr p10-15 preintegration complex, inhibits cell virion nucleus division, arrests infected cells at G2/M promotes extracellular release of viral particles; degrades CD4 in the ER; Vpu p16 integral membrane protein (protein only present in HIV-1 and SIVcpz) plasma membrane, Nef p27-p25 CD4 and class I downregulation cytoplasm Vpr homolog present in HIV-2 and Vpx p12-p16 virion some SIVs, absent in HIV-1 primarily in the Tev p28 tripartite tat-env-rev protein nucleolus/nucleus

7

Figure 1.6. HIV life cycle with descriptions of each step30.

8

The HIV life cycle consists of seven stages: (1) binding, (2) fusion, (3) reverse transcription, (4) integration, (5) transcription and translation, (6) assembly and (7) budding (Figure 1.6)31. During binding, the HIV envelope protein, gp120, binds to host cell CD4 receptors, which is followed by binding to either CCR5 or CXCR4 co-receptors.

This induces a conformational change in gp120 which allows for the release of gp41, which brings the viral membrane and the host cell membranes to close proximity and leads to membrane fusion32. Once fusion has occurred, the virus matrix and capsid are digested and viral enzymes and viral genomes are released into the cell31. When the viral single stranded RNA genomes are released into the cell they are reverse transcribed into double stranded DNA by viral reverse transcriptase31. Many of the mutations in HIV are introduced in this step as HIV reverse transcriptase has poor fidelity and is prone to errors during transcription33. After the viral DNA has been generated it associates with the HIV protein, integrase, which mediates integration of the viral DNA into the host genome34. The integrase/DNA complex preferentially targets transcriptionally active sites, which is thought to promote efficient viral gene expression after integration into the host genome34. After integration of the viral DNA into the host genome, viral transcription and translation begins. This is a multi-step process that involves host and viral proteins that is largely controlled by the viral proteins, tat and rev, which are the first proteins that the viral genome produces35. HIV transcription is mediated by tat by promoting efficient elongation of viral primary mRNA transcripts35.

After viral primary mRNA transcripts are produced, they undergo complex post transcriptional processing to produce RNA transcripts of all viral proteins35. The rev

9 protein then exports mRNA transcripts out of the nucleus into the cytoplasm where the

HIV gag protein promotes viral translation35. After translation, the viral protease cleaves viral polyproteins, specifically the gag and pol proteins, into their mature forms29,36. If this step is inhibited the new HIV particles are rendered uninfectious36. Two copies of the viral ssRNA and the mature proteins are then assembled into a new viral particle at the host cell membrane where they bud off and are released into the host36.

The pathology of HIV infection has three distinct clinical phases: acute, latent and development of AIDS37. During the acute phase of HIV infection, patients experience symptoms similar to influenza or mononucleosis with symptoms occurring one to four weeks after infection and lasting for one to two weeks38. During the acute phase there is a large increase in plasma viral load and a sharp decrease in CD4+ T-cell count37.

After the acute phase of HIV infection, the virus and T-cell counts stabilize and the infection enters the chronic or latent stage of infection37. This stage is usually asymptomatic, although lymphadenopathy and Kaposi’s sarcoma can been seen, and can last from one to twenty years before progressing37. The final stage of HIV infection is progression to AIDS which is defined by the CDC as “CD4+ T-lymphocyte count of

<200 cells/µL or CD4+ T-lymphocyte percentage of total lymphocytes of <14 or documentation of an AIDS-defining condition.”39,40 As the T-cell population plummets and the viral load increases, patients experience increases in the incidence of opportunistic infections, cancers, nervous system disorders and wasting. Patients that progress to AIDS and are untreated will generally die within three years37,41.

10

1.3 AZT Discovery and Approval

In the 1960s the National Cancer Institute (NCI) undertook efforts to identify novel drugs to treat cancer42. During this time, Jerome Horwitz at the Barbara Ann Karmanos

Cancer Institute and Wayne State University School of Medicine, first synthesized a thymidine analog which would later be called 3'-azido-3'-deoxythymidine (AZT; Figure

1.7)43. Unfortunately, this drug did not prove to be effective in treating cancer in preclinical studies, so no further testing was conducted on AZT at the time42.

Figure 1.7. Structure of AZT. Note the azide group replacing the hydroxyl group at the 3' position of the sugar moiety. Image from PubChem, Compound ID: 35370. http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=35370

11

In the 1970s, Wolfram Ostertag of the Max Planck Institute demonstrated that AZT suppressed the replication of the Friend murine leukemia virus, a retrovirus that causes cancer in murine models44. With the advent of the AIDS epidemic and the identification of the HIV virus as its cause, efforts at the NCI were undertaken to find compounds that had the ability to suppress HIV replication in vitro and would have pharmacologic properties that made them amenable for clinical studies42. Compounds that could inhibit the synthesis of nucleic acids were the focus of many of these studies, as they had already proved useful in the treatment of virus-induced cancers, notably 6- mercaptopurine for the treatment of leukemia and Non-Hodgkin’s lymphoma45. Mitsuya et.al. demonstrated in early 1985 that AZT potently suppressed HIV replication and abrogated HIV’s cytotoxic effects in vitro46.

This led the corporate sponsor of AZT, Burroughs-Wellcome, to file an

Investigational New Drug Application (IND) with the FDA to begin clinical trials of AZT in humans42. Amazingly, the FDA approved the IND in seven days42. This trial was a

Phase I efficacy and dosing trial to determine the appropriate dose of AZT for efficacious treatment of AIDS without undue toxicity47. The results from the study showed that doses of 2.5 mg/kg IV followed by 5 mg/kg orally at either 8 or 4 hour intervals showed the maximum benefit of AZT treatment, with all patients increasing T- cell counts and four out of eight patients showing reactivity to antigenic skin testing, an indication of immune system function47.

Based on the results of the Phase I AZT clinical trial, a double-blind, placebo- controlled Phase II trial to evaluate AZT efficacy and toxicity was initiated in February

1986 and enrolled 282 patients with AIDS or AIDS-related complex (ARC) which were

12 recruited into either the AZT (n=145) arm or the placebo (n=137) arm. One hundred ninety four patients were still participating when the trial was halted by its Drug Safety

Monitoring Board (DSMB) in September 198648; the study was halted due to 19 patients in the placebo arm dying versus one in the AZT arm. The AZT arm of the study showed clear benefit to survival (p < 0.001) and additional endpoints used to evaluate efficacy, regardless of subgrouping by diagnosis or CD4+ T-cell counts 48.

The results of the companion toxicity study discovered that AZT patients had increased incidence of anemia (p < 0.001) and neutropenia (p < 0.001) compared to the placebo group 49. Compared to the severity of AIDS, these toxicities were viewed as acceptable, although studies were conducted shortly thereafter that determined a reduced dose was equally efficacious and reduced toxicity50.

Three weeks after the Phase II AZT trial was halted, the FDA issued a treatment

IND, an exceedingly rare waiver that allowed patients to access AZT prior to approval by the FDA42. Burroughs-Wellcome submitted a New Drug Application (NDA) for the approval of AZT in the treatment of AIDS in late 1986 and was approved by the FDA in

March of 1987; AZT is still widely used today (Figure 1.8) 42.

13

Figure 1.8. Vials of AZT, first marketed as Retrovir. “AZT drug vials ,” HIV and AIDS 30 Years Ago, accessed March 18, 2014, Courtesy of the National Museum of American History, Smithsonian Institution, Division of Medicine and Science. http://hivaids.omeka.net/items/show/35.

1.4 Antiretroviral (ARV) Pharmacology

1.4.1 ARV Overview

With the advent of Highly Active Antiretroviral Therapy (HAART), AIDS related deaths have plummeted, changing HIV infection and AIDS from a fatal disease to a chronic condition (Figure 1.9)51.

14

Figure 1.9. Numbers of persons infected with HIV, AIDS diagnoses and deaths from 1981-200851.

Since the approval of AZT in 1987, the FDA has approved 38 drugs in seven different drug classes for the treatment of HIV (Table 1.2)52. Each of these classes, with the exception of combination therapies, targets a different and specific stage in the HIV replication cycle. Current World Health Organization (WHO) and the U.S. Department of Health and Human Services treatment guidelines recommend multiple classes of drugs be used in combination as first line and second line treatments in adults infected with HIV, with variations for special populations such as pregnant women and children53,54.

First line treatment for adults usually consists of two nucleoside reverse transcriptase inhibitors (NRTI) along with one non-nucleoside reverse transcriptase inhibitor (NNRTI); in second line therapy the NNRTI is most commonly replaced with a ritonavir boosted protease inhibitor (PI)53,54. If toxicities develop that are specific to a

15 particular drug, then a similar drug in the same class may be substituted; for example if a patient develops renal toxicity due to tenofovir (TDF) use, AZT may be substituted in the treatment regimen53,54.

In resource poor settings compared to developed countries, cost and access to treatment are the primary reasons for differences in treatment regimens. For example, in a resource poor setting, an AZT based regimen may be selected over TDF as TDF costs three times as much per non-generic pill as AZT ($33.95 per 300 mg Viread vs.

$9.10 per 300 mg Retrovir55). There also are large variations in access to medical care, societal customs and access to treatment drugs that present specific challenges to obtaining HIV treatment in undeveloped countries22.

While HAART regimens have changed HIV infection to a chronic disease state, there are limitations to HIV treatments. There are two main limitations to HAART: 1) incomplete immune reconstitution and 2) drug toxicities56. While virtually all patients that receive HAART experience viral loads <500 copies/mL, most patients do not have full immune system reconstitution57,58. This leads to inflammatory diseases and complications to HIV infection such as increased risk of cardiovascular, hepatic, renal and neurological disorders56,59. Now that HIV infected patients live life spans close to the non-infected population, patients are taking ARVs for decades59. Many of these drugs’ have side effects that are not immediately apparent, such as tenofovir and renal toxicity, and ultimately lead to treatment regimen changes or less than optimal adherence56,59,60.

16

Table 1.2. Antiretroviral (ARV) drugs approved by the FDA for the treatment of

HIV infection. Data modified from FDA52.

Time to Brand Generic Name/ Approval Drug Class Manufacturer Approval Name Abbreviation Date (Months) Emtriva emtricitabine, FTC Gilead Sciences 2-Jul-03 10 Epivir lamivudine, 3TC GlaxoSmithKline 17-Nov-95 4.4 zalcitabine, dideoxycytidine, Hivid Hoffmann-La Roche 19-Jun-92 7.6 Nucleoside ddC (no longer marketed) zidovudine, azidothymidine, Reverse Retrovir GlaxoSmithKline 19-Mar-87 3.5 Transcriptase AZT, ZDV Inhibitors Videx didanosine, dideoxyinosine, ddI Bristol Myers-Squibb 9-Oct-91 6 (NRTIs) tenofovir disoproxil fumarate, Viread Gilead 26-Oct-01 5.9 TDF Zerit stavudine, d4T Bristol Myers-Squibb 24-Jun-94 5.9 Ziagen abacavir sulfate, ABC GlaxoSmithKline 17-Dec-98 5.8 Tibotec Edurant rilpivirine 20-May-11 10 Therapeutics Non- Tibotec nucleoside Intelence etravirine, ETR 18-Jan-08 6 Reverse Therapeutics Transcriptase Rescriptor delavirdine, DLV Pfizer 4-Apr-97 8.7 Inhibitors Sustiva efavirenz, EFV Bristol Myers-Squibb 17-Sep-98 3.2 (NNRTIs) Boehringer Viramune nevirapine, NVP 21-Jun-96 3.9 Ingelheim amprenavir, APV (no longer Agenerase GlaxoSmithKline 15-Apr-99 6 marketed) Boehringer Aptivus tipranavir, TPV 22-Jun-05 6 Ingelheim Crixivan indinavir, IDV, Merck 13-Mar-96 1.4 Fortovase saquinavir (no longer marketed) Hoffmann-La Roche 7-Nov-97 5.9

Protease Invirase saquinavir mesylate, SQV Hoffmann-La Roche 6-Dec-95 3.2 Inhibitors Kaletra lopinavir and ritonavir, LPV/RTV Abbott Laboratories 15-Sep-00 3.5 (PIs) Fosamprenavir Calcium, FOS- Lexiva GlaxoSmithKline 20-Oct-03 10 APV Norvir ritonavir, RTV Abbott Laboratories 1-Mar-96 2.3 Prezista darunavir Tibotec, Inc. 23-Jun-06 6 Reyataz atazanavir sulfate, ATV Bristol-Myers Squibb 20-Jun-03 6 Agouron Viracept nelfinavir mesylate, NFV 14-Mar-97 2.6 Pharmaceuticals Fusion Hoffmann-La Roche Fuzeon enfuvirtide, T-20 13-Mar-03 6 Inhibitors & Trimeris Entry Selzentry maraviroc Pfizer 6-Aug-07 8 Inhibitors HIV Integrase Isentress raltegravir, RAL Merck & Co., Inc. 12--Oct-07 6 Inhibitors Tivicay dolutegravir GlaxoSmithKline 13-Aug-13 6

17

1.4.2 NRTI Pharmacology

NRTIs are the backbone of most antiretroviral (ARV) regimens and are most frequently prescribed in pairs with the most efficacious and least toxic combination being TDF/FTC (or 3TC)53,54. To date eight NRTIs (Table 1.2) have been approved by the FDA to treat HIV, with several other combination forms also approved and many more drugs in the development pipeline52,61. NRTIs are analogs of 2’-deoxy- nucleosides and nucleotides that all lack the 3’-OH group and require multiple intracellular phosphorylation steps for conversion to their active deoxynucleoside triphosphate (dNTP) analogs61,62. They inhibit HIV replication by competing with endogenous dNTPs for incorporation by HIV reverse transcriptase into replicating HIV proviral DNA61,62. When NRTIs are incorporated into the elongating DNA strand, they cause chain termination and stop proviral replication62.

NRTIs gain entry into cells through passive diffusion or active transport facilitated by several members of the solute-carrier (SLC) family and are actively effluxed out of cells by members of the ATP-binding cassette (ABC) transporter family61-63. Because NRTIs persist intracellularly in their active triphosphorylated form, their efficacy is correlated with intracellular concentrations of drug rather than plasma concentrations64,65. Drug pharmacokinetics may be influenced by polymorphisms in drug transporter or catabolic genes. An example of this is the ABCC4 3463A>G polymorphism (rs1751034) and

TDF, where the intracellular concentration of TDF is increased and the fraction of TDF excreted in the urine is decreased in patients with the rs1751034 polymorphism66,67.

There are two major limitations to the use of NRTIs, resistance and toxicity. The major mechanisms of resistance are exclusion and excision. Resistance due to

18 exclusion occurs when the HIV-RT gains mutations that allow it to favor endogenous dNTPs over the triphosphorylated NRTIs61. Excision based resistance occurs when the

HIV-RT gains a mutation that allows it to efficiently excise the monophosphate form of the drug (primarily AZT) from the end of the elongating viral DNA strand and allows viral elongation to continue61. Generally, resistance mutations are specific to a particular drug or dNTP analog, thus when resistance to one NRTI occurs, patients can be switched to another NRTI to maintain viral suppression53. For example, resistance mutations to abacavir (ABC), a guanosine analog, do not affect viral susceptibility to thymidine analogs such as AZT68.

The other main limitation to NRTI use is NRTI toxicity. The most common toxicities associated with all NRTIs are lipodystrophy and lactic acidosis. Some toxicities specific to a particular drug are: AZT and anemia, TDF and renal toxicity, and ABC and hypersensitivity reactions69-71. The majority of NRTI toxicities, with the exception of

ABC hypersensitivity, are related to mitochondrial damage induced by NRTIs72-75. The mechanism of NRTI induced mitochondrial toxicity is the affinity that the mitochondrial polymerase, polγ, has for NRTIs76. The mitochondria incorporate NRTIs into their genomes during replication and this leads to depletion of mtDNA, resulting in oxidative stress, mitochondrial dysfunction and ultimately cellular apoptosis77.

1.4.3 NNRTI Pharmacology

NNRTIs are also a common component of ARV regimens53,54. In 1996, nevirapine

(NVP) was the first NNRTI approved and NNRTIs were the second class of drug to be approved by the FDA for the treatment of HIV infection (Table 1.2). To date, five

19

NNRTIs have been approved for use in the treatment of HIV-1 (Table 1.2).

Interestingly, HIV-2 is naturally resistant to NNRTIs, since the site of NNRTI binding on

HIV-RT is not present in HIV-278. Like NRTIs, NNRTIs inhibit the function of HIV-RT, however, they work by allosterically binding to the active site of HIV-RT in the NNRTI binding pocket (NNIBP)61. When NNRTIs bind in the NNIBP, they force a conformational change in the active site of the HIV-RT which halts DNA polymerization61.

NNRTIs are passively absorbed into cells and undergo hepatic metabolism, primarily by CYP2B6 and CYP3A479. The hydroxylated metabolites undergo glucoronidation and may be effluxed by ABC transporters79. Because NNRTIs undergo extensive

Cytochrome P450 (CYP450) metabolism and can induce expression of CYP450 enzymes, there are numerous drug-drug interactions associated with their use54.

Multiple studies have demonstrated the effect of CYP2B6 polymorphisms on the pharmacokinetics of NVP and EFV, most notably the CYP2B6 516G>T polymorphism

(rs3745274) that has been shown repeatedly to increase plasma concentrations of both drugs in patients with the variant allele80-82. There is some controversy surrounding the effect of the ABCB1 3435C>T polymorphism (rs1045642) on the pharmacokinetics of

EFV and NVP, with one study showing an effect on EFV Cmin concentrations and several studies showing no effect79,83-87.

Like NRTIs, drug toxicity and resistance mutations are the main limitations to the use of NNRTIs in the treatment of HIV. The most common dose limiting side effects with

NVP are rash and hepatotoxicity88. Decreased risk of developing NVP hepatotoxicity has been associated with the presence of ABCB1 3435C>T (rs1045642)89,90.

20

Additionally, NVP induced rash has been associated with the HLA-DRB1*0101 allele79.

Efavirenz is mainly associated with central nervous system disorders such as insomnia and vivid dreams and these effects have been associated with increased plasma concentrations associated with CYP2B6 polymorphisms79,91.

1.4.4 PI Pharmacology

Protease inhibitors are commonly prescribed along with NRTIs for the treatment of

HIV53,54. Their utility as a monotherapy in virally suppressed patients is also being studied92. There are ten PIs approved by the FDA for the treatment of HIV infection

(Table 1.2). Protease inhibitors are peptidomimetic molecules that have a nonhydrolyzable hydroxyethylene core that cannot be cleaved by HIV protease and inhibit its function93,94. When the HIV protease is inhibited, viral polyproteins cannot be processed and the newly synthesized HIV viral particle is rendered uninfectious94. PIs are extensively metabolized by CYP450 metabolic enzymes, specifically CYP3A4, and are substrates for transporters in the SLC and ABC drug transporter families93. Many

PIs have short half-lives due to CYP3A4 metabolism93 and are commonly boosted by ritonavir, a CYP3A4 inhibitor, to allow once-daily dosing. Because many drugs are substrates of CYP3A4 and ABC transporters, there is a large risk of drug-drug interactions with ritonavir boosted PI regimens93. There have been many in vitro studies suggesting that polymorphisms in ABC transporters, OATPs, OCTs and CYP3As affect the transport and metabolism of PIs. However, many of these effects (particularly for

CYP3A4 and CYP3A5) have not been shown to have clinical implications, possibly due to ritonavir modulation and phenotypic effects93,95. One exception to this is SLCO1B1

21

521T>C (rs4149056), which has been shown to increase plasma area under the curves

(AUC) and clearance (CL) of lopinavir/ritonavir and affect viral loads in children96.

1.4.5 Integrase Inhibitor and Entry Inhibitor Pharmacology

There are two relatively new classes of drugs to treat HIV, integrase inhibitors and fusion/entry inhibitors (Table 1.2), with two drugs approved in each class. Integrase inhibitors block integration of the viral DNA complex into the host genome by inhibiting the strand transfer process mediated by HIV integrase97,98. Integrase inhibitors are primarily hepatically metabolized via glucuronidation by UGT1A199,100. There are relatively few side effects and drug-drug interactions associated with the use of integrase inhibitors due to their specificity to HIV integrase and lack of CYP450 metabolism98,101.

Fusion/entry inhibitors block the first step in HIV infection, the fusion and entry of the virus into the host cell102. Maraviroc is a CCR5 receptor agonist that prevents binding of the HIV virus to the host cell. Maraviroc is metabolized by CYP3A4, is a substrate of P- glycoprotein and has very few side effects95. Due to CYP3A4 metabolism, maraviroc does interact with other drugs using the same metabolic pathway, but these effects can be overcome with dose adjustments103.

Enfuvirtide is a fusion inhibitor that prevents the viral gp41 protein from fusing with the host cell membrane102. Enfuvirtide is a synthetic peptide that is expected to undergo catabolism into its respective amino acids104. Enfuvirtide has been relatively well tolerated; there have been sporadic reports of hypersensitivity reactions but they have

22 been unconfirmed to date105. These new classes of drugs have given hope to patients that have experienced treatment failure or drug toxicities with the older drug classes.

1.5 Pharmacogenetics of ARV Therapy

While great progress has been made in the treatment of HIV, there still remains large interindividual variation in drug response and toxicity between patients.

Pharmacogenetics is the study of how patients’ genetic backgrounds influence drug efficacy and toxicity. The field of pharmacogenetics truly began in the 1950s when the widely known effects of polymorphisms in N-acetyltransferase on isoniazid metabolism were observed106. Also during the 1950s, many studies on the heritability of drug toxicity and response were conducted using twin studies, which greatly expanded the field of pharmacogenetics106. With the advent of molecular cloning and genetic sequencing technologies, it has become possible to thoroughly investigate the effect of genetic variation on drug response and toxicity106. While much of the interindividual variation in drug response and toxicity has been explained, there still are many questions that remain to be answered. The majority of pharmacogenetic studies conducted have been in Caucasian populations, however, with the advent of large scale sequencing it has become clear that there are wide variations in the frequency of polymorphisms across differing populations107.

Pharmacogenetic studies related to ARV use have largely focused on drug metabolizing enzymes, drug transporters and drug targets79. Many clinically relevant genotype-phenotype associations have been identified and additional associations are

23 still being discovered79. Table 1.3 highlights some clinically relevant associations that have been discovered to date.

Drug metabolizing enzymes play an important role in the elimination of many drugs.

ARV drugs, specifically NNRTIs, are primarily metabolized by CYP3A4 and CYP2B6, and to a lesser extent by CYP2A6 (Table 1.3). Polymorphisms in these enzymes impact drug plasma concentrations and toxicity. Several missense mutations in

CYP2B6, 785A>G (rs2279343, Lys262Arg), 983T>C (rs28399499, Ile328Thr), and

516G>T (rs3745274, Gln172His), have been associated with increased plasma levels of

NVP and EFV as a result of a decrease in enzyme function80-82,108-111. The CYP2B6

516G>T polymorphism has also been associated with increased incidence of NVP induced hepatotoxicity112,113. The CYP3A5 6986A>G (rs776746) polymorphism causes an alternate RNA spice site which leads to a nonfunctional protein114. As a consequence of this, carriers of the variant allele of this polymorphism have a decreased NVP AUC115. There are two CYP2A6 variants that impact the metabolism of

NNRTIs, CYP2A6 47441C>T (rs28399454, Val365Met) and 6857G>T (rs8192726)

(Table 1.3). Both of these polymorphisms have been shown to increase EFV plasma concentrations, however only the 47441C>T polymorphism has been shown to increase

NVP plasma levels116. Several UGT1A1 polymorphisms are associated with increased

PI-induced hyperbilirubinemia, specifically with atazanavir and indinavir117. Additionally, the UGT2B7 802T>C (rs7439366, Tyr268His) polymorphism is associated with increased EFV plasma concentrations116.

The role that drug transporters play in the disposition of many ARVs has been well documented and polymorphisms in these genes can impact drug pharmacokinetics and

24 toxicity. ABC transporters have demonstrated the most clinically significant effects on drug pharmacokinetics and toxicity, most notably ABCC2 and ABCC4, with emerging data to support the role of ABCC10 and controversial data on the role of ABCB1 (Table

1.3). Multiple polymorphisms in the ABCC2 and ABCC4 genes are associated with increased risk of tenofovir induced renal toxicity67. The role of ABCC10 is just beginning to be recognized in the disposition of ARVs, with one polymorphism, 1791+526G>A

(rs9349256) associated with increased risk of tenofovir renal toxicity and another polymorphism, 2759T>C (rs2125739), associated with increased nevirapine plasma concentrations118,119. The role of P-gp in the disposition of ARVs remains unclear at this date and there is significant controversy surrounding the impact, if any, of the ABCB1

3435C>T (rs1045642, Ile1145Ile) and 2677G>T/A (rs2032582, Ser893Ala/Ser893Thr) on ARV pharmacokinetics and toxicity120.

Several genes that are not part of the pharmacokinetic pathways of ARVs have been demonstrated to effect ARV toxicity (Table 1.3). The most important of these is the increased risk of abacavir hypersensitivity reaction associated with the HLA-B*57:01:01 allele. The risk of abacavir hypersensitivity is greatly increased in individuals with the

HLA-B*57 allele and the Clinical Pharmacogenetics Implementation Consortium (CPIC) discourages prescribing of abacavir to patients with the allele71. Variants in several other genes are associated with increased ARV toxicity including: HTR2A, NT5C2 and

XDH (Table 1.3). Polymorphisms in HTR2A, a serotonin receptor, are associated with increased EFV central nervous system side effects and polymorphisms in NT5C2 and

XDH are associated with increased risk of didanosine induced noncirrhotic portal hypertension91,121.

25

Table 1.3. Selected polymorphisms in drug metabolizing enzymes and drug transporters associated with ARV response and toxicity.

Gene SNP Allele Drugs Function Amino Acid Change Clinical Effect efavirenz lamivudine lopinavir nelfinavir ABCB1 rs1045642 3435C>T Synonymous Ile1145Ile Controversial120 nevirapine ritonavir tenofovir zidovudine

26 efavirenz

lamivudine 2677G>A, 120 ABCB1 rs2032582 Missense Ser893Ala; Ser893Thr Controversial 2677G>T nevirapine zidovudine ABCC10 rs9349256 1791+526G>A tenofovir Intronic Risk of kidney toxicity118

ABCC10 rs2125739 2759T>C nevirapine Missense Ile948Thr TC/CC ↑ Plasma levels; EFV, NVP119 ABCC2 rs17222723 3563T>A tenofovir Missense Val1188Glu Risk of kidney toxicity67 ABCC2 rs2273697 26353G>A tenofovir Missense Val417Ile Risk of kidney toxicity67 ABCC2 rs717620 -24C>T tenofovir 5' UTR Risk of kidney toxicity67

ABCC2 rs8187710 4544G>A tenofovir Missense Cys1515Tyr Risk of kidney toxicity67

ABCC4 rs11568695 3609G>A tenofovir Synonymous Ala1203Ala Risk of kidney toxicity67 ABCC4 rs1751034 3348A>G tenofovir Synonymous Lys1116Lys Risk of kidney toxicity67

26

Gene SNP Allele Drugs Function Amino Acid Change Clinical Effect efavirenz CYP2A6 rs28399454 47441C>T Missense Val365Met Variant alleles ↑ Plasma levels116 nevirapine CYP2A6 rs8192726 6857G>T efavirenz Intronic CA/AA ↑ Plasma levels116

efavirenz 785A>G, 80 CYP2B6 rs2279343 Missense Lys262Arg AG/GG ↑ Plasma levels; EFV, NVP CYP2B6*4 nevirapine efavirenz 983T>C, part of 109,110 CYP2B6 rs28399499 Missense Ile328Thr CT/CC ↑ Plasma levels; EFV, NVP CYP2B6*18 nevirapine

516G>T, efavirenz GT/TT ↑ Plasma levels; EFV, NVP; CYP2B6 rs3745274 Missense Gln172His 80-82,111,122 CYP2B6*6 nevirapine ↑NVP Hepatotoxicity CYP3A5 rs776746 6986A>G nevirapine Acceptor ↓NVP Plasma levels115

HLA-B - *57:01:01 abacavir - - ↑ hypersensitivity71

27 HTR2A rs6313 102C>T efavirenz Intronic Ser34Ser ↑CNS Side effects91

NT5C2 rs11191561 104869531C>G didanosine Intronic ↑ Noncirrhotic portal hypertension 121

NT5C2 rs11598702 104897985T>C didanosine Intronic ↑ Noncirrhotic portal hypertension121

UGT1A1 rs4148323 211G>A indinavir Intronic Gly71Arg ↑ hyperbilirubinemia117 UGT1A1*28, atazanavir UGT1A1 rs8175347 UGT1A1*36, Not Available ↑ hyperbilirubinemia;ATZ, IDV117

UGT1A1*37 indinavir 802T>C, 116 UGT2B7 rs7439366 efavirenz Missense Tyr268His ↑EFV plasma concentration UGT2B7*2 XDH rs1429376 10410448A>C didanosine Intronic ↑ Noncirrhotic portal hypertension121

XDH rs1594160 10400856A>C didanosine Intronic ↑ Noncirrhotic portal hypertension121

27

1.6 Dissertation Aims

The aim of this dissertation is to characterize the impact of host genetics on complications due to HIV infection and ARV pharmacology and toxicity. Since the advent of HAART, many of the complications to HIV infection have not been frequently observed in the developed world. However, these complications are still a significant problem in the developing world and much is still unknown about the effect that host genetics play in the risk of developing these complications. Additionally, while numerous studies detailing environmental and genetic factors have explained a portion of the interpatient variability of ARV PK and toxicity, there is still substantial unaccounted for variability in patient populations. The main questions addressed in the studies of this dissertation are as follows:

1. Are NVP pharmacokinetics heritable in different ethnic populations?

2. What are the genetic predictors of NVP pharmacokinetics, both known and novel?

3. What are the genetic predictors of HIV induced peripheral neuropathy?

4. What are the genetic predictors of NRTI induced peripheral neuropathy?

The following studies were conducted to determine the answers to these questions.

1. To determine if NVP pharmacokinetics is heritable the relative genetic contribution

to nevirapine pharmacokinetics was characterized in African and European

Americans. Repeated dose data was used to estimate heritability of NVP exposure

and selected polymorphisms in CYP2B6 and ABCB1 were also examined.

28

2. To identify and characterize genetic predictors of NVP pharmacokinetics the relationship between nevirapine Cmin concentrations and polymorphisms in candidate genes was examined in treatment naïve HIV+ Ugandans.

3 .To identify novel genetic predictors of HIV induced peripheral neuropathy a genome-wide association study was conducted in a Ugandan HIV+ population.

Bioinformatic analyses were conducted to identify biologically plausible genetic loci associated with HIV-induced peripheral neuropathy.

4. To identify genetic predictors of NRTI induced peripheral neuropathy a genome- wide association study was conducted in a Ugandan HIV+ population. The biological plausibility of candidate genomic loci was investigated using bioinformatic tools.

29

1.7 References

1. CDC, Pneumocystis pneumonia — Los Angeles. Morbidity and Mortality Weekly Report 30, 1-3 (1981). 2. CDC, Kaposi's sarcoma and Pneumocystis pneumonia among homosexual men- --New York City and California. Morbidity and Mortality Weekly Report 30, 305- 308 (1981). 3. USAMRIID, "Medical Management of Biological Casualties Handbook," (United States Government Printing Office, Fort Detrick, Maryland, 2011). 4. K. B. Hymes, J. B. Greene, D. C. William, T. Cheung, N. S. Prose, H. Ballard, L. J. Laubenstein, Kaposi's Sarcoma in Homosexual Men - A Report of Eight Cases. The Lancet, 598-600 (1981). 5. M. S. Gottlieb, R. Schroff, H. M. Schanker, J. D. Weisman, P. T. Fan, R. A. Wolf, A. Saxon, Pneumocystis Carinii Pneumonia and Mucosal Candidiasis in Previously Healthy Homosexual Men. New England Journal of Medicine 305, 1425-1431 (1981). 6. M. D. Frederick P. Siegal, P. D. Carlos Lopez, M. D. Glenn S. Hammer, M. D. Arthur E. Brown, M. D. Stephen J. Kornfeld, M. D. Jonathan Gold, M. D. Joseph Hassett, M. D. Shalom Z. Hirschman, M. D. P. D. Charlotte Cunningham- Rundles, M. D. Bernard R. Adelsberg, M. D. David M. Parham, M. A. Marta Siegal, P. D. Susanna Cunningham-Rundles, M. D. Donald Armstrong, Severe Acquired Immunodeficiency in Male Homosexuals, Manifested by Chronic Perianal Ulcerative Herpes Simplex Legions. New England Journal of Medicine 305, 1439-1444 (1981). 7. D. T. Durack, Opportunistic Infections and Kaposi's Sarcoma in Homosexual Men. New England Journal of Medicine 305, 1465-1467 (1981). 8. H. Masur, M. A. Michelis, J. B. Greene, I. Onorato, R. A. Vande Stouwe, R. S. Holzman, G. Wormser, L. Brettman, M. Lange, H. W. Murray, S. Cunningham- Rundles, An Outbreak of Community-Acquired Pneumocyctis Carinii Pneumonia. New England Journal of Medicine 305, 1431-1437 (1981). 9. J. W. Curran, Special Report - Epidemiologic Aspects of the Current Outbreak of Kaposi's Sarcoma and Opportunistic Infections. New England Journal of Medicine 306, 248-252 (1982). 10. CDC, Current Trends Update on Acquired Immune Deficiency Syndrome (AIDS) --United States. Morbidity and Mortality Weekly Report 31, 507-508, 513-504 (1982). 11. H. W. JAFFE, K. CHOI, P. A. THOMAS, H. W. HAVERKOS, D. M. AUERBACH, M. E. GUINAN, M. F. ROGERS, T. J. SPIRA, W. W. DARROW, M. A. KRAMER, S. M. FRIEDMAN, J. M. MONROE, A. E. FRIEDMAN-KIEN, L. J.

30

LAUBENSTEIN, M. MARMOR, B. SAFAI, S. K. DRITZ, S. J. CRISPI, S. L. FANNIN, J. P. ORKWIS, A. KELTER, W. R. RUSHING, S. B. THACKER, J. W. CURRAN, National Case-Control Study of Kaposi's Sarcoma and Pneumocystis carinii Pneumonia in Homosexual Men: Part 1, Epidemiologic Results. Annals of Internal Medicine 99, 145-151 (1983). 12. M. F. Rogers, D. M. Morens, J. A. Stewart, R. M. Kaminski, T. J. Spira, P. M. Feorino, S. A. Larsen, D. P. Francis, M. Wilson, L. Kaufman, National Case- Control Study of Kaposi's Sarcoma and Pneumocystis carinii Pneumonia in Homosexual Men : Part 2, Laboratory Results. Annals of Internal Medicine 99, 151-158 (1983). 13. R. C. Gallo, P. S. Sarin, E. P. Gelmann, R. Robert-Guroff, E. Richardson, V. S. Kalyanaraman, D. Mann, G. D. Sidhu, R. E. Stahl, S. Zolla-Pazner, J. Leibowitch, M. Popovic, Isolation of Human T-Cell Leukemia Virus in Acquired Immune Deficiency Syndrome (AIDS). Science 220, 865-867 (1983). 14. F. Barré-Sinoussi, J. C. Chermann, F. Rey, M. T. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. Vézinet-Brun, C. Rouzioux, W. Rozenbaum, L. Montagnier, Isolation of a T-Lymphotropic Retrovirus from a Patient at Risk for Acquired Immune Deficiency Syndrome (AIDS). Science 220, 868-871 (1983). 15. J. A. Levy, A. D. Hoffman, S. M. Kramer, J. A. Landis, J. M. Shiambukuro, L. S. Oshiro, Isolation of Lymphocytopathic Retroviruses from San Francisco Patients with AIDS. Science 225, 840-842 (1984). 16. NIAID. (2011). 17. R. C. Gallo, M. Popovic, M. G. Sarngadharan, Isolation of proteins of htlv-iii, serological detection of antibodies to htlv-iii in sera of patients with aids and pre- aids conditions, and detection of htlv-iii infection by immuno-assays using htlv-iii and its proteins. (1984). 18. M. G. Sarngadharan, M. Popovic, L. Bruch, J. Schüpbach, R. C. Gallo, Antibodies Reactive with Human T-Lymphotropic Retroviruses ( HTLV-III ) in the Serum of Patients with AIDS. Science 224, 506-508 (1984). 19. J. N. Nkeze, N. T. Constantine, R. Y. Zhao, Eds., Laboratory Testing for HIV Infection: Advances After 28 Years, (Springer New York, 2014), pp. 81-106. 20. CDC, Provisional public health services inter-agency recommendations for screening donated blood and plasma for antibody to the virus causing acquired immunodeficiency syndrome. MMWR 1985;34:1-5. Morbidity and Mortality Weekly Report 34, 1-5 (1985). 21. C. Chhatbar, R. Mishra, A. Kumar, S. K. Singh, HIV vaccine: hopes and hurdles. Drug Discovery Today 16, 948-956 (2011)10.1016/j.drudis.2011.08.013). 22. UNAIDS, "GLOBAL REPORT: UNAIDS report on the global AIDS epidemic 2013," (2013). 23. J. Hemelaar, The origin and diversity of the HIV-1 pandemic. Trends in molecular medicine 18, 182-192 (2012)10.1016/j.molmed.2011.12.001).

31

24. P. M. Sharp, B. H. Hahn, Origins of HIV and the AIDS pandemic. Cold Spring Harbor perspectives in medicine 1, a006841 (2011)10.1101/cshperspect.a006841). 25. J. Pepin, The origins of AIDS. (Cambridge University Press, 2011). 26. N. Pant Pai, S. Shivkumar, J. M. Cajas, Does genetic diversity of HIV-1 non-B subtypes differentially impact disease progression in treatment-naive HIV-1- infected individuals? A systematic review of evidence: 1996-2010. Journal of acquired immune deficiency syndromes (1999) 59, 382-388 (2012)10.1097/QAI.0b013e31824a0628). 27. M. M. Santoro, C. F. Perno, HIV-1 Genetic Variability and Clinical Implications. ISRN microbiology 2013, 481314 (2013)10.1155/2013/481314). 28. R. A. Weiss, How does HIV cause AIDS? Science (New York, N.Y.) 260, 1273- 1279 (1993)10.1126/science.8493571). 29. T. Leitner, B. Hahn, H. M. J. Foundation, C. Kuiken, B. Foley, P. Marx, S. Wolinsky, C.-c. Lo, J. Macke, J. J. Szinger, J. Thurmond, H. Yoon, M. Zhang, HIV Sequence Compendium 2008 Editors. (2008). 30. NIAID. (2010). 31. J. A. Moss, HIV/AIDS Review. Radiologic technology 84, 247-267; quiz p.268- 270 (2013). 32. L. Vandekerckhove, C. Verhofstede, D. Vogelaers, Maraviroc: perspectives for use in antiretroviral-naive HIV-1-infected patients. The Journal of antimicrobial chemotherapy 63, 1087-1096 (2009)10.1093/jac/dkp113). 33. W.-S. Hu, S. H. Hughes, HIV-1 reverse transcription. Cold Spring Harbor perspectives in medicine 2, (2012)10.1101/cshperspect.a006882). 34. R. Craigie, F. D. Bushman, HIV DNA integration. Cold Spring Harbor perspectives in medicine 2, a006890 (2012)10.1101/cshperspect.a006890). 35. J. Karn, C. M. Stoltzfus, Regulation of HIV-1 Gene Expression. Cold Spring Harb Perspect Med 4, a006916 (2012). 36. W. I. Sundquist, H.-G. Kräusslich, HIV-1 assembly, budding, and maturation. Cold Spring Harbor perspectives in medicine 2, a006924 (2012)10.1101/cshperspect.a006924). 37. F. Epstein, G. Pantaleo, The immunopathogenesis of human immunodeficiency virus infection. New England Journal … 328, 327-335 (1993). 38. C. Chu, P. a. Selwyn, Diagnosis and initial management of acute HIV infection. American family physician 81, 1239-1244 (2010). 39. CDC, Revised Surveillance Case Definitions for HIV Infection Among Adults, Adolescents, and Children Aged <18 Months and for HIV Infection and AIDS Among Children Aged 18 Months to <13 Years --- United States, 2008. Morbidity and Mortality Weekly Report 57(RR-10), 1-12.

32

40. CDC, 1993 Revised Classification System for HIV Infection and Expanded Surveillance Case Definition for AIDS Among Adolescents and Adults. Morbidity and Mortality Weekly Report 41(RR-17), 1-19 (1993 Sept 24). 41. M. Zwahlen, M. Egger, C.-. Bern, Progression and mortality of untreated HIV- positive individuals living in resource-limited settings : Update of literature review and evidence synthesis. UNAIDS Obligation, 1-17. 42. S. Broder, The development of antiretroviral therapy and its impact on the HIV- 1/AIDS pandemic. Antiviral research 85, 1-38 (2010)10.1016/j.antiviral.2009.10.002.The). 43. J. P. Horwitz, J. Chua, M. Noel, Nucleosides. V. The Monomesylates of 1-(2'- Deoxy-β-D-lyxofuranosyl)thymine1,2. Journal of Organic Chemistry 29, 2076- 2078 (1964). 44. W. Ostertag, G. Roesler, C. J. Krieg, J. Kind, T. Cole, T. Crozier, G. Gaedicke, G. Steinheider, N. Kluge, S. Dube, Induction of endogenous virus and of thymidine kinase by bromodeoxyuridine in cell cultures transformed by Friend virus. Proceedings of the National Academy of Sciences of the United States of America 71, 4980-4985 (1974). 45. H. M. Temin, The DNA provirus hypothesis. Nobel Lecture, 1-19 (1975). 46. H. Mitsuya, K. J. Weinhold, P. A. Furman, M. H. St Clair, S. N. Lehrman, R. C. Gallo, D. Bolognesi, D. W. Barry, S. Broder, 3'-Azido-3'-deoxythymidine (BW A509U): an antiviral agent that inhibits the infectivity and cytopathic effect of human T-lymphotropic virus type III/lymphadenopathy-associated virus in vitro. Proceedings of the National Academy of Sciences of the United States of America 82, 7096-7100 (1985). 47. R. Yarchoan, R. W. Klecker, K. J. Weinhold, P. D. Markham, H. K. I. M. Lyerly, D. T. Durack, E. Gelmann, S. N. Lehrman, R. M. Blum, D. W. Barry, G. M. Shearer, J. M. Collins, C. E. Myers, M. A. Fischl, D. P. Bolognesi, S. Broder, Administration of 3'-azido-3'-deoxythymidine, an inhibitor of HTLV-III/LAV replication, to patients with AIDS or AIDS-related complex. Lancet 1, 575-580 (1986)10.1016/S0140-6736(86)92808-4). 48. M. A. Fischl, D. D. Richman, M. H. Grieco, M. S. Gottlieb, P. A. Volberding, O. L. Laskin, J. M. Leedom, J. E. Groopman, D. Mildvan, R. T. Schooley, G. G. Jackson, D. T. Durack, D. King, a. T. A. C. W. Group, The efficacy of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex A Double-Blind, Placebo-Controlled Trial. New England Journal of Medicine 317, 185-191 (1987). 49. D. D. Richman, M. A. Fischl, M. H. Grieco, M. S. Gottlieb, P. A. Volberding, O. L. Laskin, J. M. Leedom, J. E. Groopman, D. Mildvan, M. S. Hirsch, G. G. Jackson, D. T. Durack, S. Nusinoff-Lehrman, a. T. A. C. W. Group, The toxicity of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. A double-blind, placebo-controlled trial. New England Journal of Medicine 317, 192-197 (1987)10.1056/NEJM198707233170402).

33

50. M. A. Fischl, C. B. Parker, C. Pettinelli, M. Wulfsohn, M. S. Hirsch, A. C. Collier, D. Antoniskis, M. Ho, D. D. Richman, E. Fuchs, T. C. Merigan, R. C. Reichman, J. Gold, N. Steigbigel, G. S. Leoung, S. Rasheed, A. Tsiatis, a. t. A. C. T. Group*, A Randomized Controlled Trial of a Reduced Daily Dose of Zidovudine in Patients with the Acquired Immunodeficiency Symdrome. New England Journal of Medicine 323, 1009-1014 (1990). 51. CDC, Thirty Years of HIV - 1981-2011. Morbidity and Mortality Weekly Report 60, 1-40 (2011). 52. FDA. vol. 2014. 53. World Health Organization, "Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection," (2013). 54. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. (2013). 55. C. Knox, V. Law, T. Jewison, P. Liu, S. Ly, A. Frolkis, A. Pon, K. Banco, C. Mak, V. Neveu, Y. Djoumbou, R. Eisner, A. C. Guo, D. S. Wishart, DrugBank 3.0: a comprehensive resource for 'omics' research on drugs. Nucleic Acids Research 39, D1035-1041 (2011)10.1093/nar/gkq1126). 56. S. G. Deeks, S. R. Lewin, D. V. Havlir, The end of AIDS: HIV infection as a chronic disease. Lancet 382, 1525-1533 (2013)10.1016/S0140-6736(13)61809- 7). 57. C. F. Kelley, C. M. R. Kitchen, P. W. Hunt, B. Rodriguez, F. M. Hecht, M. Kitahata, H. M. Crane, J. Willig, M. Mugavero, M. Saag, J. N. Martin, S. G. Deeks, Incomplete peripheral CD4+ cell count restoration in HIV-infected patients receiving long-term antiretroviral treatment. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 48, 787-794 (2009)10.1086/597093). 58. K. N. Althoff, K. Buchacz, H. I. Hall, J. Zhang, D. B. Hanna, P. Rebeiro, S. J. Gange, R. D. Moore, M. M. Kitahata, K. A. Gebo, J. Martin, A. C. Justice, M. A. Horberg, R. S. Hogg, T. R. Sterling, A. Cescon, M. B. Klein, J. E. Thorne, H. M. Crane, M. J. Mugavero, S. Napravnik, G. D. Kirk, U.S. Trends in Antiretroviral Therapy Use, HIV RNA Plasma Viral Loads, and CD4 T-Lymphocyte Cell Counts Among HIV-Infected Persons, 2000 to 2008. Annals of Internal Medicine 157, 325-335 (2012). 59. A. C. Justice, HIV and aging: time for a new paradigm. Current HIV/AIDS reports 7, 69-76 (2010)10.1007/s11904-010-0041-9). 60. A. E. Zimmermann, T. Pizzoferrato, J. Bedford, A. Morris, R. Hoffman, G. Braden, Tenofovir-associated acute and chronic kidney disease: a case of multiple drug interactions. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 42, 283-290 (2006)10.1086/499048). 61. S. G. Sarafianos, B. Marchand, K. Das, D. M. Himmel, M. a. Parniak, S. H. Hughes, E. Arnold, Structure and function of HIV-1 reverse transcriptase:

34

molecular mechanisms of polymerization and inhibition. Journal of molecular biology 385, 693-713 (2009)10.1016/j.jmb.2008.10.071). 62. T. Cihlar, A. S. Ray, Nucleoside and nucleotide HIV reverse transcriptase inhibitors: 25 years after zidovudine. Antiviral research 85, 39-58 (2010)10.1016/j.antiviral.2009.09.014). 63. O. Kis, K. Robillard, G. N. Y. Chan, R. Bendayan, The complexities of antiretroviral drug-drug interactions: role of ABC and SLC transporters. Trends in pharmacological sciences 31, 22-35 (2010)10.1016/j.tips.2009.10.001). 64. C. Bazzoli, V. Jullien, C. Le Tiec, E. Rey, F. Mentré, A.-m. Taburet, C. L. Tiec, F. Mentre, Intracellular Pharmacokinetics of Antiretroviral Drugs in HIV-Infected Patients, and their Correlation with Drug Action. Clinical Pharmacokinetics 49, 17-45 (2010)10.2165/11318110-000000000-00000). 65. P. L. Anderson, T. N. Kakuda, K. A. Lichtenstein, The cellular pharmacology of nucleoside- and nucleotide-analogue reverse-transcriptase inhibitors and its relationship to clinical toxicities. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 38, 743-753 (2004)10.1086/381678). 66. J. J. Kiser, M. L. Carten, C. L. Aquilante, P. L. Anderson, P. Wolfe, T. M. King, T. Delahunty, L. R. Bushman, C. V. Fletcher, The Effect of Lopinavir / Ritonavir on the Renal Clearance of Tenofovir in HIV-infected Patients. Clinical Pharmacology & Therapeutics 83, 5-8 (2008). 67. J. J. Kiser, C. L. Aquilante, P. L. Anderson, T. M. King, M. L. Carten, C. V. Fletcher, Clinical and genetic determinants of intracellular tenofovir diphosphate concentrations in HIV-infected patients. J. Acquir. Immune Defic. Syndr 47, 298- 303 (2008). 68. F. Fitzgerald, M. Penazzato, D. Gibb, Development of antiretroviral resistance in children with HIV in low- and middle-income countries. The Journal of infectious diseases 207 Suppl S85-92 (2013)10.1093/infdis/jit115). 69. E. R. Scruggs, A. J. Dirks Naylor, Mechanisms of zidovudine-induced mitochondrial toxicity and myopathy. Pharmacology 82, 83-88 (2008)10.1159/000134943). 70. J. J. Kohler, S. H. Hosseini, A. Hoying-Brandt, E. Green, D. M. Johnson, R. Russ, D. Tran, C. M. Raper, R. Santoianni, W. Lewis, Tenofovir renal toxicity targets mitochondria of renal proximal tubules. Laboratory investigation; a journal of technical methods and pathology 89, 513-519 (2009)10.1038/labinvest.2009.14). 71. M. a. Martin, T. E. Klein, B. J. Dong, M. Pirmohamed, D. W. Haas, D. L. Kroetz, Clinical pharmacogenetics implementation consortium guidelines for HLA-B genotype and abacavir dosing. Clinical pharmacology and therapeutics 91, 734- 738 (2012)10.1038/clpt.2011.355).

35

72. W. Lewis, B. J. Day, W. C. Copeland, Mitochondrial toxicity of NRTI antiviral drugs: an integrated cellular perspective. Nature reviews. Drug discovery 2, 812- 822 (2003)10.1038/nrd1201). 73. A. Maagaard, D. Kvale, Mitochondrial toxicity in HIV-infected patients both off and on antiretroviral treatment: a continuum or distinct underlying mechanisms? The Journal of antimicrobial chemotherapy 64, 901-909 (2009)10.1093/jac/dkp316). 74. C. M. Bailey, K. S. Anderson, A mechanistic view of human mitochondrial DNA polymerase gamma: providing insight into drug toxicity and mitochondrial disease. Biochimica et biophysica acta 1804, 1213-1222 (2010)10.1016/j.bbapap.2010.01.007). 75. B. Liang, Æ. D. J. Kleinhenz, Æ. E. R. Walp, Æ. S. Dikalov, D. P. Jones, Æ. R. F. Schinazi, Æ. R. L. Sutliff, E. R. Kline, L. Bassit, B. I. Hernandez-Santiago, M. a. Detorio, D. J. Kleinhenz, E. R. Walp, S. Dikalov, R. F. Schinazi, R. L. Sutliff, Long-term exposure to AZT, but not d4T, increases endothelial cell oxidative stress and mitochondrial dysfunction. Cardiovascular toxicology 9, 1-12 (2009)10.1007/s12012-008-9029-8). 76. W. Lewis, Pharmacogenomics, toxicogenomics, and DNA polymerase gamma. The Journal of infectious diseases 195, 1399-1401 (2007)10.1086/513879). 77. J. J. Kohler, W. Lewis, A Brief Overview of Mechanisms of Mitochondrial Toxicity From NRTIs. Journal of Clinical Virology 172, 166-172 (2007)10.1002/em). 78. I. Boehringer Ingelheim Pharamaceuticals. (Ridgefield, CT, 2005). 79. V. Tozzi, Pharmacogenetics of antiretrovirals. Antiviral research 85, 190-200 (2010)10.1016/j.antiviral.2009.09.001). 80. M. Rotger, S. Colombo, H. Furrer, G. Bleiber, T. Buclin, B. L. Lee, O. Keiser, J. Biollaz, L. Décosterd, A. Telenti, Influence of CYP2B6 polymorphism on plasma and intracellular concentrations and toxicity of efavirenz and nevirapine in HIV- infected patients. Pharmacogenetics and genomics 15, 1-5 (2005). 81. A. Saitoh, E. Sarles, E. Capparelli, F. Aweeka, A. Kovacs, S. K. Burchett, A. Wiznia, S. Nachman, T. Fenton, S. A. Spector, CYP2B6 genetic variants are associated with nevirapine pharmacokinetics and clinical response in HIV-1- infected children. AIDS (London, England) 21, 2191-2199 (2007)10.1097/QAD.0b013e3282ef9695). 82. M. Arab-Alameddine, J. Di Iulio, T. Buclin, M. Rotger, R. Lubomirov, M. Cavassini, A. Fayet, L. A. Décosterd, C. B. Eap, J. Biollaz, A. Telenti, C. Csajka, Pharmacogenetics-based population pharmacokinetic analysis of efavirenz in HIV-1-infected individuals. Clinical pharmacology and therapeutics 85, 485-494 (2009)10.1038/clpt.2008.271). 83. J. Fellay, C. Marzolini, E. R. Meaden, D. J. Back, T. Buclin, J. P. Chave, L. a. Decosterd, H. Furrer, M. Opravil, G. Pantaleo, D. Retelska, L. Ruiz, A. H. Schinkel, P. Vernazza, C. B. Eap, A. Telenti, Response to antiretroviral treatment in HIV-1-infected individuals with allelic variants of the multidrug resistance

36

transporter 1: a pharmacogenetics study. Lancet 359, 30-36 (2002)10.1016/S0140-6736(02)07276-8). 84. S. G. Heil, M. E. V. D. Ende, P. W. Schenk, I. V. D. Heiden, J. Lindemans, D. Burger, R. H. N. V. Schaik, Associations Between ABCB1, CYP2A6, CYP2B6, CYP2D6, and CYP3A5 Alleles in Relation to Efavirenz and Nevirapine Pharmacokinetics in HIV-Infected Individuals. Therapeutic drug monitoring 34, 153-159 (2012). 85. L. Elens, B. Vandercam, J. C. Yombi, D. Lison, P. Wallemacq, V. Haufroid, Influence of host genetic factors on efavirenz plasma and intracellular pharmacokinetics in HIV-1-infected patients. Pharmacogenomics 11, 1223-1234 (2010); published online EpubSep (10.2217/pgs.10.94). 86. H. J. Ribaudo, H. Liu, M. Schwab, E. Schaeffeler, M. Eichelbaum, A. A. Motsinger-Reif, M. D. Ritchie, U. M. Zanger, E. P. Acosta, G. D. Morse, R. M. Gulick, G. K. Robbins, D. Clifford, D. W. Haas, Effect of CYP2B6, ABCB1, and CYP3A5 polymorphisms on efavirenz pharmacokinetics and treatment response: an AIDS Clinical Trials Group study. J Infect Dis 202, 717-722 (2010); published online EpubSep (10.1086/655470). 87. R. Winzer, P. Langmann, M. Zilly, F. Tollmann, J. Schubert, H. Klinker, B. Weissbrich, No influence of the P-glycoprotein polymorphisms MDR1 G2677T/A and C3435T on the virological and immunological response in treatment naïve HIV-positive patients. Ann Clin Microbiol Antimicrob 4, 3 (2005)10.1186/1476- 0711-4-3). 88. L. F. Chen, J. Hoy, S. R. Lewin, Ten years of highly active antiretroviral therapy for HIV infection. The Medical journal of Australia 186, 146-151 (2007). 89. M. D. Ritchie, D. W. Haas, A. A. Motsinger, J. P. Donahue, H. Erdem, S. Raffanti, P. Rebeiro, A. L. George, R. B. Kim, J. L. Haines, T. R. Sterling, Drug transporter and metabolizing enzyme gene variants and nonnucleoside reverse-transcriptase inhibitor hepatotoxicity. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 43, 779-782 (2006)10.1086/507101). 90. D. W. Haas, J. a. Bartlett, J. W. Andersen, I. Sanne, G. R. Wilkinson, J. Hinkle, F. Rousseau, C. D. Ingram, A. Shaw, M. M. Lederman, R. B. Kim, Pharmacogenetics of nevirapine-associated hepatotoxicity: an Adult AIDS Clinical Trials Group collaboration. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 43, 783-786 (2006)10.1086/507097). 91. A. S. Martín, S. C. Figueroa, R. C. Guerrero, L. P. Hurtado, A. D.-G. Hurlé, Á. C. Álvarez, Impact of pharmacogenetics on CNS side effects related to efavirenz R esearch A rticle. Pharmacogenomics 14, 1167-1178 (2013). 92. M. Estébanez, J. R. Arribas, Protease inhibitor monotherapy: what is its role? Current HIV/AIDS reports 9, 179-185 (2012)10.1007/s11904-012-0112-1).

37

93. L. Griffin, P. Annaert, K. I. M. L. R. Brouwer, Influence of Drug Transport Proteins on the Pharmacokinetics and Drug Interactions of HIV Protease Inhibitors. Journal of Pharmaceutical Sciences 100, 3636-3654 (2011)10.1002/jps). 94. E. De Clercq, The nucleoside reverse transcriptase inhibitors, nonnucleoside reverse transcriptase inhibitors, and protease inhibitors in the treatment of HIV infections (AIDS). Advances in pharmacology (San Diego, Calif.) 67, 317-358 (2013)10.1016/B978-0-12-405880-4.00009-3). 95. V. Michaud, T. Bar-Magen, J. Turgeon, D. Flockhart, Z. Desta, M. a. Wainberg, The dual role of pharmacogenetics in HIV treatment: mutations and polymorphisms regulating antiretroviral drug resistance and disposition. Pharmacological Reviews 64, 803-833 (2012)10.1124/pr.111.005553). 96. N. Y. Rakhmanina, M. Neely, R. H. N. V. Schaik, H. G.-. Dressman, K. D. Williams, S. J. Soldin, J. N. van den Anker, CYP3A5, ABCB1 and SLCO1B1 Polymorphisms and Pharmacokinetics and Virologic Outcome of Lopinavir/Ritonavir in HIV-infected Children. Therapeutic Drug Monitoring 33, 417-424 (2012)10.1097/FTD.0b013e318225384f.CYP3A5). 97. L. Dickinson, S. Khoo, D. Back, Pharmacokinetics and drug-drug interactions of antiretrovirals: an update. Antiviral research 85, 176-189 (2010)10.1016/j.antiviral.2009.07.017). 98. K. C. Brown, S. Paul, A. D. M. Kashuba, Drug interactions with new and investigational antiretrovirals. Clinical Pharmacokinetics 48, 211-241 (2009). 99. K. Kassahun, I. Mcintosh, D. Cui, D. Hreniuk, S. Merschman, K. Lasseter, N. Azrolan, M. Iwamoto, J. A. Wagner, L. A. Wenning, Metabolism and Disposition in Humans of Raltegravir ( MK-0518 ), an Anti-AIDS Drug Targeting the Human Immunodeficiency Virus 1 Integrase Enzyme ABSTRACT :. Pharmacology 35, 1657-1663 (2007)10.1124/dmd.107.016196.due). 100. GlaxoSmithKline. (Research Triangle Park, NC, 2013), pp. 1-37. 101. L. a. Wenning, a. S. Petry, J. T. Kost, B. Jin, S. a. Breidinger, I. DeLepeleire, E. J. Carlini, S. Young, T. Rushmore, F. Wagner, N. M. Lunde, F. Bieberdorf, H. Greenberg, J. a. Stone, J. a. Wagner, M. Iwamoto, Pharmacokinetics of raltegravir in individuals with UGT1A1 polymorphisms. Clinical pharmacology and therapeutics 85, 623-627 (2009)10.1038/clpt.2009.12). 102. T. J. Henrich, D. R. Kuritzkes, HIV-1 entry inhibitors: recent development and clinical use. Current opinion in virology 3, 51-57 (2013)10.1016/j.coviro.2012.12.002). 103. S. Abel, T. M. Jenkins, L. a. Whitlock, C. E. Ridgway, G. J. Muirhead, Effects of CYP3A4 inducers with and without CYP3A4 inhibitors on the pharmacokinetics of maraviroc in healthy volunteers. British journal of clinical pharmacology 65 Suppl 1, 38-46 (2008)10.1111/j.1365-2125.2008.03134.x). 104. Genentech. (2013), pp. 1-18.

38

105. C. M. Davis, W. T. Shearer, Diagnosis and management of HIV drug hypersensitivity. The Journal of allergy and clinical immunology 121, 826- 832.e825 (2008)10.1016/j.jaci.2007.10.021). 106. U. A. Meyer, Pharmacogenetics - five decades of therapeutic lessons from genetic diversity. Nature reviews genetics 5, 669-676 (2004). 107. L. Clarke, X. Zheng-Bradley, R. Smith, E. Kulesha, C. Xiao, I. Toneva, B. Vaughan, D. Preuss, R. Leinonen, M. Shumway, S. Sherry, P. Flicek, The 1000 Genomes Project: data management and community access. Nature Methods 9, 459-462 (2012)10.1038/nmeth.1974). 108. M. Rotger, Predictive value of known and novel alleles of CYP2B6 for efavirenz plasma concentrations in HIV-infected individuals. Clin. Pharmacol, (2006). 109. D. W. Haas, T. Gebretsadik, G. Mayo, U. N. Menon, E. P. Acosta, A. Shintani, M. Floyd, C. M. Stein, G. R. Wilkinson, Associations between CYP2B6 polymorphisms and pharmacokinetics after a single dose of nevirapine or efavirenz in African americans. J Infect Dis 199, 872-880 (2009); published online EpubMar 15 (10.1086/597125). 110. C. Wyen, H. Hendra, M. Vogel, C. Hoffmann, H. Knechten, N. H. Brockmeyer, J. R. Bogner, J. Rockstroh, S. Esser, H. Jaeger, T. Harrer, S. Mauss, J. van Lunzen, N. Skoetz, A. Jetter, C. Groneuer, G. Fätkenheuer, S. H. Khoo, D. Egan, D. J. Back, A. Owen, Impact of CYP2B6 983T>C polymorphism on non- nucleoside reverse transcriptase inhibitor plasma concentrations in HIV-infected patients. The Journal of antimicrobial chemotherapy 61, 914-918 (2008)10.1093/jac/dkn029). 111. S. K. Gupta, S. L. Rosenkranz, Y. S. Cramer, S. L. Koletar, L. a. Szczech, V. Amorosa, S. D. Hall, The pharmacokinetics and pharmacogenomics of efavirenz and lopinavir/ritonavir in HIV-infected persons requiring hemodialysis. AIDS (London, England) 22, 1919-1927 (2008)10.1097/QAD.0b013e32830e011f). 112. C. Ciccacci, P. Borgiani, S. Ceffa, E. Sirianni, M. C. Marazzi, A. M. D. Altan, G. Paturzo, P. Bramanti, G. Novelli, L. Palombi, Nevirapine-induced hepatotoxicity and pharmacogenetics: a retrospective study in a population from Mozambique. Pharmacogenomics 11, 23-31 (2010). 113. C. Ciccacci, D. Di Fusco, M. C. Marazzi, I. Zimba, F. Erba, G. Novelli, L. Palombi, P. Borgiani, G. Liotta, Association between CYP2B6 polymorphisms and Nevirapine-induced SJS/TEN: a pharmacogenetics study. European journal of clinical pharmacology, (2013)10.1007/s00228-013-1549-x). 114. P. Kuehl, J. Zhang, Y. Lin, J. Lamba, M. Assem, J. Schuetz, P. B. Watkins, A. Daly, S. A. Wrighton, S. D. Hall, P. Maurel, M. Relling, C. Brimer, K. Yasuda, R. Venkataramanan, S. Strom, K. Thummel, M. S. Boguski, E. Schuetz, Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression. Nat Genet 27, 383-391 (2001); published online EpubApr (10.1038/86882).

39

115. K. C. Brown, M. C. Hosseinipour, J. M. Hoskins, R. K. Thirumaran, H.-C. Tien, R. Weigel, J. Tauzie, I. Shumba, J. K. Lamba, E. G. Schuetz, H. L. McLeod, A. D. M. Kashuba, A. H. Corbett, in Pharmacogenomics. (2012), vol. 13, pp. 113-121. 116. A. Kwara, M. Lartey, K. W. Sagoe, E. Kenu, M. H. Court, CYP2B6, CYP2A6 and UGT2B7 genetic polymorphisms are predictors of efavirenz mid-dose concentration in HIV-infected patients. AIDS 23, 2101-2106 (2009); published online EpubOct 23 (10.1097/QAD.0b013e3283319908). 117. M. Rotger, P. Taffe, G. Bleiber, H. F. Gunthard, H. Furrer, P. Vernazza, H. Drechsler, E. Bernasconi, M. Rickenbach, A. Telenti, Gilbert syndrome and the development of antiretroviral therapy-associated hyperbilirubinemia. The Journal of infectious diseases 192, 1381-1386 (2005)10.1086/466531). 118. S. P. Pushpakom, N. J. Liptrott, S. Rodriguez-Novoa, P. Labarga, V. Soriano, M. Albalater, E. Hopper-Borge, S. Bonora, G. Di Perri, D. J. Back, S. Khoo, M. Pirmohamed, A. Owen, Genetic variants of ABCC10, a novel tenofovir transporter, are associated with kidney tubular dysfunction. J Infect Dis 204, 145- 153 (2011); published online EpubJul 1 (10.1093/infdis/jir215). 119. N. J. Liptrott, S. Pushpakom, C. Wyen, G. Fatkenheuer, C. Hoffmann, S. Mauss, H. Knechten, N. H. Brockmeyer, E. Hopper-Borge, M. Siccardi, D. J. Back, S. H. Khoo, M. Pirmohamed, A. Owen, Association of ABCC10 polymorphisms with nevirapine plasma concentrations in the German Competence Network for HIV/AIDS. Pharmacogenet Genomics 22, 10-19 (2012); published online EpubJan (10.1097/FPC.0b013e32834dd82e). 120. A. Owen, M. Pirmohamed, S. H. Khoo, D. J. Back, Pharmacogenetics of HIV therapy. Pharmacogenetics and genomics 16, 693-703 (2006)10.1097/01.fpc.0000236338.41799.57). 121. E. Vispo, M. Cevik, J. K. Rockstroh, P. Barreiro, M. Nelson, A. Scourfield, C. Boesecke, J. C. Wasmuth, V. Soriano, Genetic determinants of idiopathic noncirrhotic portal hypertension in HIV-infected patients. Clin Infect Dis 56, 1117- 1122 (2013); published online EpubApr (10.1093/cid/cit001). 122. C. Ciccacci, D. Di Fusco, M. C. Marazzi, I. Zimba, F. Erba, G. Novelli, L. Palombi, P. Borgiani, G. Liotta, Association between CYP2B6 polymorphisms and Nevirapine-induced SJS/TEN: a pharmacogenetics study. European journal of clinical pharmacology 69, 1909-1916 (2013); published online EpubNov (10.1007/s00228-013-1549-x).

40

Chapter 2

2. Measuring the Overall Genetic Component of Nevirapine Pharmacokinetics and the Role of Selected Polymorphisms: Towards Addressing the Missing Heritability in Pharmacogenetic Phenotypes?*

2.1 Abstract

Nevirapine is an important component of highly active antiretroviral therapy used in the treatment of human immunodeficiency virus infection. There is considerable variation in the pharmacokinetics of nevirapine and this variation can impact the efficacy and toxicity of nevirapine. While some of this variation can be attributed to environmental factors, the degree to which heritability influences nevirapine pharmacokinetics is unknown. This study aims to estimate how much variation in nevirapine pharmacokinetics is due to genetic factors and to investigate the contribution of selected polymorphisms to this variability. Two doses of immediate-release nevirapine were administered to European (n=11) and African American (n=6) subjects recruited from the Research in Access to Care in the Homeless (REACH) cohort. A repeated-dose drug administration (RDA) method was used to determine the relative genetic contribution (rGC) to variability in nevirapine AUC0-6h. Nevirapine plasma levels were quantified using LC-MS/MS. Patients were also genotyped for selected polymorphisms in candidate genes that may influence nevirapine pharmacokinetics. A significant rGC for nevirapine AUC0-6h was found in Europeans (p = 0.02) and African

Americans (p = 0.01). A trend towards higher nevirapine AUC0-6h for the CYP2B6

* The text of this dissertation chapter is a reprint of the material as it appears in Pharmacogenetics1. The co-author Dr. Deanna Kroetz listed in this publication directed and supervised the research that forms the basis for the chapter.

41

516TT (rs3745274; Q172H) genotype was observed in European Americans (p = 0.19).

This study demonstrates that there is a significant genetic component to variability in nevirapine pharmacokinetics. While genetic variants such as CYP2B6 polymorphisms may contribute to some of this variation, these data suggest that there are additional genetic factors that influence nevirapine pharmacokinetics.

42

2.2 Introduction

The importance of understanding the role of genetics in variation in pharmacokinetics and pharmacodynamics has been recognized since the 1950s2-5.

Twin studies have historically been used to determine the heritability of genetic diseases and traits; these studies have also been used to determine the heritability of pharmacodynamic and pharmacokinetic parameters6. While twin studies are a useful technique to determine genetic contributions to pharmacokinetic variation, it can be impractical to use twins in pharmacogenetic studies due to difficulty in recruitment and the need to expose them to drugs. A statistical technique that was specifically developed to address this issue is the repeated drug administration (RDA) method, which uses repeated administrations of a drug to the same individuals to compare the within subject and between-subject variation in pharmacokinetic parameters7. This comparison can be used to quantify the relative genetic contribution to variations in pharmacokinetic parameters of a drug. While the RDA method is useful in determining whether pharmacokinetic or pharmacodynamic parameters of a drug have strong genetic components, it may vary with the route of administration or patient population studied7. Additionally, while one pharmacokinetic parameter for a given drug may have a strong relative genetic component, other parameters may not due to the genes involved in the absorption, metabolism and excretion of a drug8. Repeated drug administration has successfully been employed to characterize the genetic contribution to variability in pharmacokinetic parameters of several drugs, including erythromycin, midazolam and metformin9,10. However, the genetic contribution to pharmacokinetic parameter variability for many drugs is still unknown.

43

Nevirapine is a non-nucleoside reverse transcriptase inhibitor widely used as a component of antiretroviral therapy in the treatment of human immunodeficiency virus

(HIV)11. Nevirapine exhibits considerable variability in its pharmacokinetic properties; however, only part of this variability can be explained by environmental factors and concomitant conditions12. Variation in nevirapine pharmacokinetics can lead to reduced efficacy, increased viral resistance and increased toxicities13. Nevirapine is metabolized to its primary metabolite 3-hydroxynevirapine by CYP2B614. The CYP2B6 516G>T

(rs3745274) and CYP2B6 983T>C (rs28399499) variant alleles have a significant effect on nevirapine plasma levels and the CYP2B6 516T allele has also been associated with increased recovery of CD4+ T-cell populations in pediatric patients following initiation of nevirapine-containing antiretroviral therapy 15-17. Additionally, ABCB1 3435C>T

(rs1045642) has been associated with protection against nevirapine-induced hepatotoxicity and increased nevirapine concentrations in cerebral spinal fluid18,19.

Despite evidence that nevirapine pharmacokinetics are influenced by specific polymorphisms, there has not been a study conducted to quantify the relative genetic contribution to variability in nevirapine pharmacokinetics.

This study uses the repeated drug administration method to quantify the relative genetic contribution to variability in nevirapine pharmacokinetics. A significant relative genetic contribution to variation in nevirapine exposure was shown in two ethnic populations. The contribution of CYP2B6 516G>T and ABCB1 3435C>T to variability in nevirapine pharmacokinetics was also investigated.

44

2.3 Materials and Methods

2.3.1 Study Design and Subjects:

Subjects were recruited from the Research in Access to Care in the Homeless

(REACH) cohort as previously described20. Study participants are marginally housed

HIV positive individuals living in San Francisco. Seventeen patients were recruited to participate in a pharmacokinetic study where subjects receiving 200 mg nevirapine twice daily consented to pharmacokinetic blood sampling. All subjects were on therapy at least four months and were concomitantly receiving two nucleoside reverse transcriptase inhibitors. Subjects were presumed to have reached steady state concentrations. Blood samples were drawn at 0, 1, 2, 3 and 6 hr post-dose. The time between time courses varied from 13 days to 173 days. European American (n=11) and

African American (n=6) patients were included in this study. Ethnicity was self-reported and verified through genotyping of 112 ancestry informative markers and analysis using the STRUCTURE program21-23. The study was approved by the University of California

San Francisco Institutional Review Board and all subjects provided written informed consent prior to participation.

2.3.2 Nevirapine Quantification:

Plasma was prepared from blood samples by centrifugation and stored at -80°C until analysis. Nevirapine was extracted using Oasis HLB SPE columns (Waters Corp.,

Milford, MA) and plasma concentrations were quantified by LC/MS/MS analysis as described by Mistri et. al24. Briefly, each 0.5 mL plasma aliquot was heated for 1.5 hr at

56°C to inactivate HIV-1 virus and then spiked with 25 l of 20 M metaxolone (Toronto

45

Research Chemicals, Toronto, Ontario) in methanol, which served as an internal standard. SPE columns were equilibrated with 1 mL methanol followed by 1 mL distilled water. Samples were then loaded on the column and washed with 1 mL of 2 mM ammonium acetate followed by 1 mL of water. Samples were eluted in 1 mL mobile phase (80:20 acetonitrile:water, 0.1% acetic acid) and a 5 μl aliquot was injected onto a

5 μm Hypersil BDS C18 column, 50 x 4.6 μm (Thermo Fisher Scientific, Waltham, MA).

The flow rate into the API4000 mass spectrometer (AbSciex, Framingham, MA) was 0.2 mL/min and nevirapine retention time was 1.7 min. The parent ion (267.2 m/z, amu) and product ion (226.2 m/z, amu) were monitored at Q1 and Q3, respectively.

Nevirapine standard curves were linear from 50 - 5000 ng/mL (r2 > 0.9). Assay accuracy was between 100.3% and 112.9% relative standard deviation. Assay precision ranged from 8.2 – 18.5% CV.

2.3.3 Genotyping:

Genomic DNA was extracted from whole blood samples. Genotyping of polymorphisms of interest (CYP2B6 516G>T and ABCB1 3435C>T) was accomplished using the ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster

City, CA). TaqMan assays were used to genotype CYP2B6 516G>T (rs3745274, Assay

ID: C___7817765_60) and ABCB1 3435C>T (rs1045642, Assay ID: C___7586657_20).

Genotypes were called with ABI Sequence Detection System software (version 2.1;

Applied Biosystems, Foster City, CA).

46

2.3.4 Calculation of Pharmacokinetic Parameters:

25 Due to the long half-life of nevirapine (45 hr), only AUC0-6h was calculated . AUC0-6h was calculated for each dose administration using the trapezoidal rule.

2.3.5 Calculation of Relative Genetic Component:

The genetic contribution to the variability in nevirapine AUC0-6h was assessed with a modified ANOVA formula for estimating the relative genetic component or rGC and 95% confidence intervals proposed by Kalow et.al.26:

2 2 2 rGC = (SDb -SDw )/SDb

which can be rearranged as

2 2 rGC = 1-(1/F) where F = SDb /SDw

Upper and Lower Confidence intervals can be calculated using:

Lower 95% confidence interval = Fobserved/F0.025,b.d.f,w.d.f

Upper 95% confidence interval = Fobserved*F0.025,b.d.f,w.d.f

2 where rGC represents the estimated relative genetic component, SDb is the between

2 subjects variation, SDw is the within subject variation, b.d.f is the between subjects degrees of freedom, w.d.f. is the within subject degrees of freedom and F0.025 is the tabulated F statistic at the 2.5% significance level at the appropriate degrees of freedom. Due to well characterized differences in allele frequency and linkage disequilibrium patterns, European Americans (n=11) and African Americans (n=6) were analyzed separately in this study.

47

2.3.6 Statistical Methods:

Statistical significance for genetic contribution to AUC0-6h variability was calculated using an F-test, α=0.05, to determine if the inter- and intra-individual variation was significantly different. One-way ANOVA, α=0.05, was used to determine significance for the effect of genetic polymorphisms on AUC0-6h values. All other calculations of p- values were obtained using two-sided t-tests or one-way ANOVA as appropriate27.

Calculations were performed using R and Microsoft Excel28. All figures were produced in Prism Version 5.01 (GraphPad Software Inc., San Diego, CA).

2.4 ResultsEthnicity does not play a role in nevirapine AUC0-6h variability

Since there are well characterized differences in the genetic structure and linkage disequilibrium patterns in different ethnic populations, a statistical analysis to examine any overall differences in nevirapine AUC0-6h between African and European Americans was conducted. A total of 17 subjects were included in this study, 11 European

Americans and six African Americans (Table 2.1). Median ages and concomitant medications were similar in the two ethnic groups, while the African American group had a higher proportion of females than the European American group.

Analysis of nevirapine plasma concentrations indicated very little intrasubject variability in concentrations during the six hours following drug administration, consistent with the long terminal half-life of this drug (see Figures 2.1A and 2.1B). In contrast, there is considerable variation in nevirapine concentrations between individuals; three individuals in the African American and two in the European American groups never reach plasma concentrations above the minimum effective concentration

48

29 (MEC) for nevirapine of 3000 μg/L . Average AUC0-6h did not differ between the two visits, although there was significant interpatient variability in these values (Table 2.1).

For example, the mean AUC0-6h was 22.5 mg nevirapine/L*hr (SEM = 3.81 mg nevirapine/L*hr) and 18.3 mg nevirapine/L*hr (SEM = 2.69 mg nevirapine/L*hr) for

European and African Americans, respectively. There was not a significant difference in

AUC0-6h between ethnicities (t-test, p = 0.45).

Table 2.1 Patient demographics and relative genetic contribution (rGC) to

nevirapine AUC0-6h European African

Americans Americans Sample Size n 11 6 Male (%) 4 (36) 1 (17) Gender Female (%) 7 (64) 5 (83) Median 45 49 Age (years) Range 29 - 57 33 - 74 2 SDw 2.39 5.34 Nevirapine AUC0-6h 1 (mg/L*hr) 2 SDb 24.9 54.7 0.904 0.902 r 2 (95% CI) GC (0.64-0.97) (0.42-0.98) Estimated Relative Genetic Component F 10.4 10.2 p 0.02 0.01 1 2 2 SDw is within individual variation and SDb is between subject variation. 2 Estimated relative genetic component

49

A

B

Figure 2.1. Interindividual variation in nevirapine plasma levels. Plasma concentrations of A) African American subjects and B) European American subjects 0-6 hours after nevirapine dose administration. Each line represents one individual.

50

2.4.2 Age and sex do not play a role in the variability of nevirapine AUC0-6h

To ensure further analyses were not confounded by demographic factors, the effects of age and sex on nevirapine AUC0-6h were examined by linear regression and t-tests,

2 respectively. Age had no effect on nevirapine AUC0-6h with an r of 0.04. Males tended to have slightly lower AUC0-6h (16.2 mg nevirapine/L*hr, SEM = 37.0 mg nevirapine/L*hr) than females (23.0 mg nevirapine/L*hr, SEM=24.2 mg nevirapine/L*hr) however, this difference was not statistically significant (p = 0.14).

2.4.3 There is a genetic contribution to variation in nevirapine AUC0-6h

The relative genetic contribution to nevirapine pharmacokinetics was calculated using the repeated drug administration method described previously7,9. The between-

2 subject (SDb ) variation in AUC0-6h was about 10-fold greater than the within subject

2 variation (SDw ) in both ethnic groups (Table 2.1). The calculated rGC and upper and lower 95% confidence intervals for the European Americans and African Americans was

0.902 (0.64 - 0.97) and 0.904 (0.42 - 0.98), respectively. F-tests indicate there is a significant genetic contribution to the variability in AUC0-6h in both Europeans (p = 0.02) and African Americans (p = 0.01).

2.4.4 CYP2B6 516G>T may influence nevirapine AUC0-6h

Considering the evidence for a significant genetic contribution to the variability in nevirapine exposure, polymorphisms in candidate genes implicated in the metabolism and transport of nevirapine were tested for association with nevirapine pharmacokinetics. In African Americans, there is a trend for increased plasma nevirapine levels in individuals carrying the CYP2B6 516G>T allele or the ABCB1

51

3435C>T allele (Figures 2.2A and 2.3A); however, the sample sizes are too small for formal statistical analysis (Table 2.2). A similar trend was observed for the CYP2B6

516G>T allele in European Americans, but these differences did not reach statistical significance (Figure 2.2B and Table 2.2). There was no indication of an association between the ABCB1 3435C>T polymorphism and nevirapine pharmacokinetics in

European Americans (Figure 2.3B and Table 2.2).

52

Figure 2.2. Nevirapine plasma concentrations following a single oral dose. A 200 mg dose of nevirapine was administered to A) African Americans and B) European Americans and concentrations were measured over six hours. The concentrations (mean ± SEM) are stratified by CYP2B6 516G>T genotype: circles GG, squares GT and triangles TT.

53

Figure 2.3. Nevirapine plasma concentrations following a single oral dose. A 200 mg dose of nevirapine was administered to A) African Americans and B) European Americans and concentrations were measured over six hours. The concentrations (mean ± SEM) are stratified by ABCB1 3435C>T genotype: circles CC and squares CT.

54

Table 2.2. The effect of ethnicity and genotype on nevirapine exposure Nevirapine AUC Ethnicity n 0-6h p (mg/L*h)1 African American 6 18.3 ± 3.81 0.45 European American 12 22.5 ± 2.69

African Americans

GG 1 4.23

GT 5 21.2 ± 5.63 ND

TT 0 - CYP2B6 516G>T European Americans

GG 4 20.5 ± 3.16

GT 5 20.3 ± 2.82 0.19

TT 3 28.8 ± 3.64

African Americans

CC 2 7.31 ± 7.97 0.17 CT 4 23.8 ± 12.8 ABCB1 3435C>T European Americans

CC 6 22.0 ± 3.03 0.96 CT 5 22.2 ± 3.32 1 Mean ± SEM

55

2.5 Discussion

While there have been many candidate gene association studies of nevirapine pharmacokinetics, this is the first study to determine the overall relative genetic influence on nevirapine exposure. A significant relative genetic contribution to the variability in nevirapine pharmacokinetics was demonstrated in European and African

Americans. This supports previous findings that have implicated polymorphisms in drug metabolism and transport genes in nevirapine pharmacokinetic variability and toxicity16-

18. A trend consistent with previous studies of elevated plasma concentrations in subjects homozygous for the CYP2B6 516G>T allele was also observed15,17.

Variability in nevirapine pharmacokinetics and toxicity has been observed since its approval for the treatment of HIV. Many candidate gene studies have confirmed that a portion of pharmacokinetic variability is due to polymorphisms in CYP2B616,30,31.

However, the variation in pharmacokinetics due to genetic versus environmental factors has never been examined. The current study demonstrates that there is a significant genetic component to nevirapine pharmacokinetics in African and European Americans.

While the population examined here is small, one advantage of the RDA method is the ability to use small populations to estimate relative genetic components of drugs8. In our European population, we have the required number of subjects to estimate a 95%

8 Lower Confidence limit of ~0.65 for an rGC of 0.9 . This suggests that interindividual variation in nevirapine drug levels could be reduced through knowledge of a patient’s genetic background. The importance of this is reflected in the observation that several patients did not reach the MEC of nevirapine. The RDA method has been successfully employed to identify drugs whose renal clearance has a strong genetic component and

56 could also be used to identify antiretroviral drugs that are good candidates for pharmacogenomics research10. Employing the RDA method in pharmacogenomic research could lead to decreased efficacy against HIV and increased viral resistance to nevirapine and other antiretroviral drugs.

To further investigate the influence of genetics on nevirapine pharmacokinetics, two candidate polymorphisms were selected for study. A trend was observed towards elevated AUC0-6h of nevirapine in both European and African Americans homozygous for the CYP2B6 516G>T polymorphism. This polymorphism is associated with a slight decrease in hepatic protein expression and function, therefore increases in AUC0-6h are expected32. While the results in European Americans did not reach statistical significance, the analysis was limited by a small sample size and may have been confounded by unidentified environmental factors. The trend observed is consistent with other published work, which supports the need for a larger study population16,17,33.

No association of ABCB1 3435C>T with nevirapine exposure was observed in our study. The effect of this polymorphism on nevirapine pharmacokinetics remains controversial, with many studies not showing an effect on nevirapine plasma

15-17,31,34 pharmacokinetics . AUC0-6h may not be the most appropriate pharmacokinetic parameter to observe the effects of these polymorphisms; however, due to the long half- life of nevirapine, it was not possible to calculate other pharmacokinetic parameters such as half-life or oral clearance.

57

2.6 Conclusions

The current study demonstrates that there is a significant relative genetic component to nevirapine pharmacokinetics. While there are genetic variants such as CYP2B6 polymorphisms that have been attributed to some of this variation16,17,33, this study suggests that there may be additional genetic factors that influence nevirapine pharmacokinetics. This study supports additional research to discover novel genetic factors influencing nevirapine variability. Furthermore, the RDA method could also be used to study endpoints of antiretroviral drugs other than pharmacokinetic and pharmacodynamic parameters, such as metabolomic endpoints35. Additional knowledge of genetic factors that affect nevirapine pharmacokinetics may help increase the efficacy of nevirapine in the treatment of HIV and lead to less viral resistance over time.

58

2.7 References

1. J. E. Micheli, L. W. Chinn, S. B. Shugarts, A. Patel, J. N. Martin, D. R. Bangsberg, D. L. Kroetz, Measuring the overall genetic component of nevirapine pharmacokinetics and the role of selected polymorphisms: towards addressing the missing heritability in pharmacogenetic phenotypes? Pharmacogenetics and genomics, 1-6 (2013)10.1097/FPC.0b013e32836533a5). 2. A. G. Motulsky, Drug Reactions, Enzymes and Biochemical Genetics. Council on Drugs 165, 835-837 (1957). 3. P. E. Carson, C. L. Flanagan, C. E. Ickes, A. S. Alving, S.-t. Relationships, Enzymatic Deficiency in Primaquine-Sensitive Erythrocytes. Science 124, 484- 485 (1956). 4. W. Kalow, GENETIC FACTORS IN RELATION TO DRUGS. Annual review of pharmacology 5, 9-26 (1964). 5. E. S. Wesell, Pharmacogenetic Perspectives Gained From Twin and Family Studies. Pharmac. Ther. 41, 535-552 (1989). 6. W. Kalow, L. Endrenyi, B. Tang, Repeat administration of drugs as a means to assess the genetic component in pharmacological variability. Pharmacology 58, 281-284 (1999). 7. V. Ozdemir, W. Kalow, L. Tothfalusi, L. Bertilsson, L. Endrenyi, J. E. Graham, Multigenic Control of Drug Response and Regulatory Decision-Making in Pharmacogenomics: The Need for an Upper-Bound Estimate of Genetic Contributions. Current Pharmacogenomics 3, 53-71 (2005)10.2174/1570160053175027). 8. V. Ozdemir, W. Kalow, B. K. Tang, a. D. Paterson, S. E. Walker, L. Endrenyi, a. D. Kashuba, Evaluation of the genetic component of variability in CYP3A4 activity: a repeated drug administration method. Pharmacogenetics 10, 373-388 (2000). 9. M. K. Leabman, K. M. Giacomini, Estimating the contribution of genes and environment to variation in renal drug clearance. Pharmacogenetics 13, 581-584 (2003)10.1097/01.fpc.0000054111.14659.2b). 10. P. o. A. G. f. A. a. Adolescents, Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Department of Health and Human Services, (2013). 11. M. Gandhi, L. Z. Benet, P. Bacchetti, A. Kalinowski, K. Anastos, A. R. Wolfe, M. Young, M. Cohen, H. Minkoff, S. J. Gange, R. M. Greenblatt, Nonnucleoside reverse transcriptase inhibitor pharmacokinetics in a large unselected cohort of HIV-infected women. Journal of acquired immune deficiency syndromes (1999) 50, 482-491 (2009). 12. C. L. Cooper, R. P. G. van Heeswijk, Once-daily nevirapine dosing: a pharmacokinetics, efficacy and safety review. HIV medicine 8, 1-7 (2007)10.1111/j.1468-1293.2007.00426.x).

59

13. P. Riska, M. Lamson, T. Macgregor, J. Sabo, S. Hattox, J. Pav, J. Keirns, DISPOSITION AND BIOTRANSFORMATION OF THE ANTIRETROVIRAL DRUG NEVIRAPINE IN HUMANS ABSTRACT :. Pharmacology 27, (1999). 14. T. Mahungu, C. Smith, F. Turner, D. Egan, M. Youle, M. Johnson, S. Khoo, D. Back, a. Owen, Cytochrome P450 2B6 516G-->T is associated with plasma concentrations of nevirapine at both 200 mg twice daily and 400 mg once daily in an ethnically diverse population. HIV medicine 10, 310-317 (2009)10.1111/j.1468-1293.2008.00689.x). 15. S. R. Penzak, G. Kabuye, P. Mugyenyi, F. Mbamanya, V. Natarajan, R. M. Alfaro, C. Kityo, E. Formentini, H. Masur, Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients in Uganda. HIV medicine 8, 86-91 (2007)10.1111/j.1468-1293.2007.00432.x). 16. David W. W. Haas, T. Gebretsadik, G. Mayo, Usha N. N. Menon, Edward P. P. Acosta, A. Shintani, M. Floyd, C. M. M. Stein, Grant R. R. Wilkinson, Associations between CYP2B6 polymorphisms and pharmacokinetics after a single dose of nevirapine or efavirenz in African americans. The Journal of infectious diseases 199, 872-880 (2009)10.1086/597125). 17. C. Ciccacci, P. Borgiani, S. Ceffa, E. Sirianni, M. C. Marazzi, A. M. D. Altan, G. Paturzo, P. Bramanti, G. Novelli, L. Palombi, Nevirapine-induced hepatotoxicity and pharmacogenetics: a retrospective study in a population from Mozambique. Pharmacogenomics 11, 23-31 (2010). 18. L. Varatharajan, S. a. Thomas, The transport of anti-HIV drugs across blood- CNS interfaces: summary of current knowledge and recommendations for further research. Antiviral research 82, A99-109 (2009)10.1016/j.antiviral.2008.12.013). 19. A. R. Moss, J. a. Hahn, S. Perry, E. D. Charlebois, D. Guzman, R. a. Clark, D. R. Bangsberg, Adherence to highly active antiretroviral therapy in the homeless population in San Francisco: a prospective study. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 39, 1190-1198 (2004)10.1086/424008). 20. H.-J. Tsai, S. Choudhry, M. Naqvi, W. Rodriguez-Cintron, E. G. Burchard, E. Ziv, Comparison of three methods to estimate genetic ancestry and control for stratification in genetic association studies among admixed populations. Human Genetics 118, 424-433 (2005). 21. J. K. Pritchard, M. Stephens, P. Donnelly, Inference of population structure using multilocus genotype data. Genetics 155, 945-959 (2000). 22. D. Falush, M. Stephens, J. K. Pritchard, Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164, 1567-1587 (2003). 23. H. N. Mistri, A. G. Jangid, A. Pudage, N. Gomes, M. Sanyal, P. Shrivastav, High throughput LC-MS/MS method for simultaneous quantification of lamivudine, stavudine and nevirapine in human plasma. Journal of chromatography. B,

60

Analytical technologies in the biomedical and life sciences 853, 320-332 (2007)10.1016/j.jchromb.2007.03.047). 24. I. Boehringer Ingelheim Pharamaceuticals. (Ridgefield, CT, 2005). 25. W. Kalow, B. K. Tang, L. Endrenyi, Hypothesis: Comparisons of inter- and intra- individual variations can substitute for twin studies in drug research. Pharmacogenetics 8, 283-289 (1998). 26. W. N. Venables, B. D. Ripley, Modern Applied Statistics with S-Plus. Statistics and Computing, 462 (1994). 27. R. C. Team, R. F. f. S. Computing, R: A Language and Environment for Statistical Computing. (Vienna, Austria, 2012). 28. T. E. M. S. D. Vries-sluijs, J. P. Dieleman, D. Arts, A. D. R. Huitema, J. H. Beijnen, M. Schutten, M. E. V. Der, Concentrations Predict Virological Failure in an Unselected HIV-1-Infected Population. Clinical pharmacokinetics 42, 599-605 (2003). 29. G. Ramachandran, K. Ramesh, A. K. Hemanth Kumar, I. Jagan, M. Vasantha, C. Padmapriyadarsini, G. Narendran, S. Rajasekaran, S. Swaminathan, Association of high T allele frequency of CYP2B6 G516T polymorphism among ethnic south Indian HIV-infected patients with elevated plasma efavirenz and nevirapine. The Journal of antimicrobial chemotherapy 63, 841-843 (2009)10.1093/jac/dkp033). 30. A. Saitoh, E. Sarles, E. Capparelli, F. Aweeka, A. Kovacs, S. K. Burchett, A. Wiznia, S. Nachman, T. Fenton, S. A. Spector, CYP2B6 genetic variants are associated with nevirapine pharmacokinetics and clinical response in HIV-1- infected children. AIDS (London, England) 21, 2191-2199 (2007)10.1097/QAD.0b013e3282ef9695). 31. T. Lang, K. Klein, J. Fischer, A. K. Nüssler, P. Neuhaus, U. Hofmann, M. Eichelbaum, M. Schwab, U. M. Zanger, Extensive genetic polymorphism in the human CYP2B6 gene with impact on expression and function in human liver. Pharmacogenetics 11, 399-415 (2001). 32. S. G. Heil, M. E. V. D. Ende, P. W. Schenk, I. V. D. Heiden, J. Lindemans, D. Burger, R. H. N. V. Schaik, Associations Between ABCB1, CYP2A6, CYP2B6, CYP2D6, and CYP3A5 Alleles in Relation to Efavirenz and Nevirapine Pharmacokinetics in HIV-Infected Individuals. Therapeutic drug monitoring 34, 153-159 (2012). 33. J. Chen, J. Sun, Q. Ma, Y. Yao, Z. Wang, L. Zhang, L. Li, F. Sun, H. Lu, CYP2B6 polymorphism and nonnucleoside reverse transcriptase inhibitor plasma concentrations in Chinese HIV-infected patients. Therapeutic drug monitoring 32, 573-578 (2010)10.1097/FTD.0b013e3181ea953c). 34. M. a. Ghannoum, P. K. Mukherjee, R. J. Jurevic, M. Retuerto, R. E. Brown, M. Sikaroodi, J. Webster-Cyriaque, P. M. Gillevet, Metabolomics reveals differential levels of oral metabolites in HIV-infected patients: toward novel diagnostic targets. Omics : a journal of integrative biology 17, 5-15 (2013)10.1089/omi.2011.0035).

61

Chapter 3

3. CYP2B6 and ABCC10 Polymorphisms Influence Nevirapine Exposure in HIV+ Ugandans

3.1 Abstract:

Nevirapine is an important component of highly active antiretroviral therapy used in the treatment of human immunodeficiency virus infection. There is considerable variation in the pharmacokinetics of nevirapine and this variation can impact its efficacy and toxicity. Some of the variation in nevirapine pharmacokinetics can be attributed to polymorphisms in CYP2B6, but other genes may also play a role in this variation. This study examined the effect of polymorphisms in CYP2B6, CYP2C19, CYP2C9, ABCC10,

NR1L2, CYP2D6, CYP3A4 and CYP3A5 on nevirapine pharmacokinetics. Patients on antiretroviral therapy regimens containing nevirapine were recruited from the Uganda

AIDS Rural Treatment Outcomes cohort. Plasma samples were taken before patients received their daily dose of nevirapine. Nevirapine trough levels were quantified using

LC-MS/MS. DNA samples were genotyped for selected polymorphisms in candidate genes that may influence nevirapine pharmacokinetics. The combined effect of multiple polymorphisms on nevirapine exposure was also explored. Several polymorphisms significantly influence nevirapine trough levels. CYP2B6 516G>T (rs3745274; p = 0.03),

CYP2B6 983T>C (rs28399499; p = 0.003) and ABCC10 rs2125739 (p = 0.001) were associated with higher nevirapine trough levels. Additionally, the number of variants in the composite of CYP2B6 516/983 (p = 0.0002) was associated with increases in nevirapine plasma concentrations. Finally, the variant load in a CYP2B6/ABCC10 composite (p = 2.5x10-6) was strongly associated with an increase in nevirapine

62 concentration. This study demonstrates the importance of CYP2B6 and ABCC10 in nevirapine pharmacokinetics. The results also support consideration of the combined effects of multiple polymorphisms on nevirapine trough levels.

63

3.2 Introduction:

Sub-Saharan Africa accounts for 69% of worldwide HIV infections and it is important to understand the consequences of common polymorphisms in African populations on the pharmacokinetics of commonly used antiretroviral (ARV) drugs1. ARV treatment for

HIV infections in sub-Saharan Africa generally consists of two nucleoside reverse transcriptase inhibitors (NRTIs) and one non-nucleotide reverse transcriptase inhibitor

(NNRTI), usually in a fixed dose combination form2. Nevirapine (NVP) is a NNRTI commonly used in sub-Saharan Africa to treat HIV and is also the recommended first line treatment to prevent mother to child transmission of HIV3,4. While NVP is an effective treatment for HIV, it has variable pharmacokinetic properties that can affect its efficacy and toxicity5,6. There have been many studies exploring what genetic and environmental factors contribute to NVP pharmacokinetics, however, there still is unaccounted for variability in NVP disposition.

NVP pharmacokinetics has been shown to be heritable and several genetic variants contribute to this variability7-11. NVP is hepatically metabolized primarily by CYP3A4 and

CYP2B6 with the latter being a major metabolic enzyme upon autoinduction12,13.

CYP2B6 516G>T (rs3745274) and CYP2B6 983T>C (rs28399499) influence the pharmacokinetics of NVP and efavirenz9-11,14,15. NVP is also a substrate of the efflux pump MRP7 (ABCC10) and polymorphisms in ABCC10 are associated with the pharmacokinetics of NVP16.

CYP2B6 516G>T is a missense mutation that causes an amino acid change from a glutamine to a histidine at the 172 position in the protein. This leads to decreased expression and function of CYP2B617,18. CYP2B6 983T>C is also a missense mutation

64 that causes an amino acid change from an isoleucine to a tyrosine at the 328 position in the protein. The CYP2B6 328Tyr variant is associated with decreased expression and function18. CYP2B6 516G>T and CYP2B6 983T>C have minor allele frequencies of

42% and 12%, respectively, in the African Yoruban (YRI) population. The consequences of both CYP2B6 polymorphisms have been extensively studied for the pharmacokinetics of efavirenz, but have been less studied for their impact on NVP pharmacokinetics8,9,19-21. ABCC10 rs2125739 has not been well studied and has only been implicated in NVP pharmacokinetics in Caucasian populations16. A survey of the literature was used to identify genes important in the biotransformation and transport of

NVP (Figure 3.1). Genes selected for inclusion in the exploratory analysis were

CYP2B6, CYP2C19, CYP2C9, ABCC10, NR1L2, CYP2D6, CYP3A4 and

CYP3A512,13,22. CYP2C19, CYP2C9, CYP2D6 and CYP3A5 were included in the candidate gene study because of their role in NVP metabolism12. PXR, encoded by

NR1L2, was included because of its ability to regulate basal expression levels of

CYP3A4/523.

This study examines the impact of individual polymorphisms in candidate genes on

NVP trough plasma concentrations. It also considers the combined effect of selected polymorphisms as a predictor of NVP exposure.

3.3 Materials and Methods:

3.3.1 Study Design and Patients:

Patients were recruited from the Uganda AIDS Rural Treatment Outcomes cohort.

Study participants are treatment naïve HIV+ patients living in Mbarra, Uganda. Patients

65 enrolled in the study receive a treatment regimen consisting of two NRTI and one

NNRTI. Patients routinely have blood drawn to monitor CD4+ cell counts. Plasma was obtained from 121 patients receiving NVP as a component of their ARV therapy. All samples were collected prior to the administration of the morning dose as a measure of

Cmin. The study was approved by the University of California San Francisco Institutional

Review Board and all subjects provided written informed consent prior to participation.

In the event that there are cultural literacy reasons why a signature is not appropriate, participants are allowed to mark consent forms with a thumbprint.

3.3.2 Nevirapine Quantification:

Blood samples were centrifuged for plasma isolation and samples were stored at -

80°C until analysis. NVP was extracted using Oasis HLB SPE columns (Waters Corp.,

Milford, MA) and plasma concentrations were quantified by LC/MS/MS analysis as described by Mistri et al24. Briefly, each 0.5 mL plasma aliquot was heated for 1.5 hrs at

56°C to inactivate HIV-1 virus and then spiked with 25 µl of 20 µM metaxolone (Toronto

Research Chemicals, Toronto, Ontario) in methanol, which served as an internal standard. SPE columns were equilibrated with 1 mL methanol followed by 1 mL distilled water. Samples were then loaded on the column and washed with 1 mL of 2 mM ammonium acetate followed by 1 mL of water. Samples were eluted in 1 mL mobile phase (80:20 acetonitrile:water, 0.1% acetic acid) and a 5 μl aliquot was injected onto a

5 μm Hypersil BDS C18 column, 50 x 4.6 μm (Thermo Fisher Scientific, Waltham, MA).

The flow rate into the API4000 mass spectrometer (AbSciex, Framingham, MA) was 0.2 mL/min and NVP retention time was 1.7 min. The parent ion (267.2 m/z, amu) and product ion (226.2 m/z, amu) were monitored at Q1 and Q3, respectively. NVP standard

66 curves were linear from 50 - 5000 ng/mL (r2 > 0.9). Assay accuracy was between

100.3% and 112.9% relative standard deviation. Assay precision ranged from 8.2 –

18.5% CV.

3.3.3 Genotyping:

Genomic DNA was extracted from saliva samples. Genotyping of CYP2B6 516G>T was accomplished using a Taqman assay (Applied Biosystems, Assay ID:

C___7817765_60) and the ABI Prism 7900HT Sequence detection system (Applied

Biosystems). Genotypes were called with ABI Sequence Detection System software

(version 2.1; Applied Biosystems). All other genotypes were determined using an

Illumina OmniExpress Bead Chip as outlined in Chapter 4.3.2. Genotypes were called with Illumina Bead Studio Software. A Χ2 test of observed vs. expected genotypes was used to determine Hardy-Weinberg equilibrium.

67

Figure 3.1. PharmGKB.org NVP pharmacokinetic pathway23. PharmGKB©. Permission to reproduce this has been granted by Stanford University and PharmGKB. Stars indicate genes that have genetic variants which have been shown to have a significant impact on NVP pharmacokinetics.

68

3.3.4 Statistical Methods:

3.3.4.1 Univariate Analysis

Nevirapine Cmin values were log transformed for statistical analyses to better approximate a normal distribution. SNPs within ±10 kb of the selected genes were included in the analysis. Only SNPs that had minor allele frequencies greater than 5% in the study population were included in the analysis. Univariate analyses using linear regression or ANOVA, α=0.05, were used to determine significance for the effect of continuous or categorical demographic covariates on Cmin values. Linear regression was used to estimate the effect of imputed dosage genotype on NVP Cmin values with a significance level of α = 0.05. SNPs in high linkage disequilibrium (r2 > 0.8) were collapsed with the highest p-value SNP retained. Scatter plots showing log NVP concentration vs. genotype were constructed and examined for Hardy-Weinberg

Equilibrium (HWE) and were removed from further analysis if a Chi-Squared test comparing the observed versus expected genotypes was p < 0.001. To reduce the amount of false positives generated by a sparse number of samples in genotype groups, genotypes were converted from dosage form to allelic genotypes. SNPs that had less than five subjects in the variant homozygote category were combined with the heterozygote category. SNPs that were significant at the level of 0.05 in the linear regression analysis were then reanalyzed using the allelic genotype calls and an

ANOVA test, α=0.05, to confirm the validity of the results. CYP2B6 516G>T was directly genotyped using a Taqman assay and only an allelic ANOVA test was performed for this SNP. Bonferroni multiple testing corrections were performed for each

69 gene separately based on the number of haplotype blocks per gene. SNPs with ANOVA adjusted p values < 0.05 were included in the multivariate analyses (Figure 3..2).

Bioinformatic analyses were conducted on all SNPs with ANOVA adjusted p values

≤ 0.05 to determine the potential for functional or regulatory consequences and linkage disequilibrium patterns were examined using SCANdb (University of Chicago, http://www.scandb.org/), RegulomeDB and HaploReg25,26. Analyses were performed using R27.

Figure 3.2. Flow chart of statistical analyses. 70

3.3.4.2 Multivariate Analysis: Multiple linear regression was used to estimate the effect of demographic and genetic variables on NVP Cmin values in multivariate analyses. Only demographic variables with a p ≤ 0.2 and SNPs with an adjusted p value of ≤ 0.05 in the univariate analysis were included in multivariate analyses. Analyses were performed using R27.

Figures were produced in R and Prism Version 5.01 (GraphPad Software Inc., San

Diego, CA).

3.4 Results

3.4.1 Characteristics of Study Participants and Analysis of the Effect of

Demographic Characteristics on NVP Cmin:

A total of 121 subjects were included in the analysis. Table 3.1 describes the demographics of the patient population. The patient population was 76% female and the median age of study participants was 35 years (range, 21 - 75 years). The patients were extremely adherent to their treatment regimens, with a median adherence of 100%

(range, 50 - 100%). Since all but five patients were > 80% adherent and because adherence data were unavailable for all patients in the study group it was not included in further statistical analyses. Patients were on a standard regimen of two NRTIs along with NVP. The majority (77%) of patients had an NRTI regimen containing zidovudine

(AZT) and lamivudine (3TC).

Univariate analyses were performed to assess the impact of demographic covariates on NVP Cmin. Age (linear regression, p = 0.102) and gender (t-test, p = 0.163) had p ≤

71

0.2, and were included in multivariate analyses. Concomitant NRTI therapy (t-test, p =

0.53) did not have significant effects on NVP Cmin.

Table 3.1. Patient Demographics and Effect on NVP Cmin

Characteristic p1 Male 29 (24) 0.163 Gender n (%) Female 92 (76) Median 35 0.102 Age (years) Range 21-75 Median 100 - Adherence (%) Range 50-100 AZT+3TC 98 (77) 0.53 Concomitant NRTIs n (%) D4T+3TC 23 (23) 1 p denotes the effect of each variable on NVP Cmin concentration. Age and adherence are linear regression p values. Gender and Concomitant NRTIs are ANOVA p values.

3.4.2 Several polymorphisms are associated with NVP Cmin:

The initial univariate analysis examined 1804 SNPs for association with NVP Cmin plasma concentrations. Of those SNPs, 55 were associated (unadjusted p < 0.05) with

NVP Cmin plasma concentrations. There were 33 unique loci after filtering for LD (Table

3.2). After recoding the genotypes from dosage to allelic format and reexamining the

SNPs using ANOVA, four SNPs had adjusted p values < 0.05 (Table 3.2).

72

Table 3.2. Linkage Disequilibrium Filtered Top Variants Associated with NVP Cmin Upper Lower ANOVA Adjusted SNP CHR Gene MAF Beta p 1 95% CI 95% CI p p rs1987236 19 CYP2B6 0.27 -0.16 -0.061 -0.259 0.002 NS rs2253635 10 CYP2C9 0.24 0.13 0.209 0.051 0.004 NS rs4917623 10 CYP2C19 0.22 0.13 0.209 0.051 0.004 0.050 0.050 rs7903917 10 CYP2C9 0.25 0.14 0.239 0.041 0.005 NS rs4388808 10 CYP2C19 0.18 0.12 0.219 0.021 0.008 0.028 0.028 rs2475376 10 CYP2C9 0.20 0.12 0.219 0.021 0.011 Removed rs60618718 19 CYP2B6 0.29 0.19 0.348 0.032 0.014 NS rs11188082 10 CYP2C19 0.36 -0.13 -0.031 -0.229 0.015 NS rs2096069 10 CYP2C9 0.27 0.11 0.209 0.011 0.015 Removed rs954356 3 NR1I2 0.09 -0.14 -0.021 -0.259 0.016 0.040 0.360 rs73933721 19 CYP2B6 0.28 0.19 0.348 0.032 0.016 NS rs9332209 10 CYP2C9 0.07 0.35 0.647 0.053 0.019 NS rs2125739 6 ABCC10 0.20 0.11 0.209 0.011 0.020 0.002 0.009 rs2224566 10 CYP2C19 0.22 0.20 0.358 0.042 0.020 NS rs8105382 19 CYP2B6 0.28 0.17 0.328 0.012 0.026 NS rs11188091 10 CYP2C19 0.22 0.18 0.338 0.022 0.027 NS rs12721652 19 CYP2B6 0.30 0.17 0.328 0.012 0.027 NS rs75249760 10 CYP2C19 0.08 0.30 0.577 0.023 0.028 NS rs2472682 3 NR1I2 0.16 0.10 0.199 0.001 0.029 NS rs74699808 19 CYP2B6 0.20 -0.20 -0.022 -0.378 0.029 NS rs4918690 10 CYP2C19 0.23 0.17 0.328 0.012 0.031 NS rs79460985 3 NR1I2 0.09 -0.25 -0.012 -0.488 0.035 0.040 0.360 rs28399499 19 CYP2B6 0.08 0.23 0.448 0.012 0.038 0.003 0.008 rs6956305 7 CYP3A5 0.11 0.12 0.239 0.001 0.039 NS rs73933726 19 CYP2B6 0.08 0.23 0.448 0.012 0.039 NS rs12721612 3 NR1I2 0.07 0.14 0.279 0.001 0.040 0.030 0.270 rs61557439 19 CYP2B6 0.32 0.16 0.318 0.002 0.041 NS rs11528090 10 CYP2C19 0.19 0.18 0.358 0.002 0.047 Removed rs16974790 19 CYP2B6 0.31 0.15 0.308 -0.008 0.047 NS rs57830676 7 CYP3A5 0.14 0.22 0.438 0.002 0.048 NS rs60549239 19 CYP2B6 0.11 -0.28 -0.003 -0.557 0.048 NS rs4688035 3 NR1I2 0.28 0.09 0.189 -0.009 0.048 Removed

1 NS not significant; adjusted p value > 0.05. Removed denotes SNPs that were removed after failing HWE testing as outlined in the Statistical methods section.

73

3.4.3 Polymorphisms in CYP2B6, ABCC10 and CYP2C19 have significant effects

on NVP Cmin

Table 3.3 describes the effect of the four significant SNPs on NVP Cmin. CYP2B6

516G>T was included in the table due to its unadjusted significance and its reported effects on NNRTI concentration, however it is not known to be a predictor of NVP concentrations in African populations.

Table 3.3. Relationship Between Genotypic Variants and NVP Cmin

NVP C (mg/L) p SNP Genotype min n (%) p Mean ± St. Dev. adjusted* TT 2.4 ± 0.9 81 (67) ABCC10 TC 3.1 ± 1.4 36 (30) 0.002 0.009a rs2125739 CC 3.5 ± 1.3 4 (3) GG 2.3 ± 1.2 49 (41) CYP2B6 516G>T GT 2.6 ± 1.0 54 (45) 0.03 0.09 rs3745274 TT 3.3 ± 1.4 16 (14) CYP2B6 983T>C TT 2.5 ± 1.1 98 (81) 0.003 0.008a rs28399499 TC/CC 3.2 ± 1.3 23 (18) CYP2C19 TT 2.4 ± 0.9 74 0.03 0.03a rs4917623 TC/CC 2.9 ± 1.4 47 CYP2C19 AA 2.4 ± 0.9 83 0.05 0.05a rs4388808 AG/GG 2.9 ± 1.5 38 * P values adjusted for the number of haplotype blocks per gene: ABCC10 = 4; CYP2B6 = 3; CYP2C19 = 1. a Variables that have significant adjusted p values

CYP2B6 983T>C (adjusted p = 0.008, Figure 3.) and ABCC10 rs2125739 (adjusted p = 0.009, Figure 3.4) genotype were associated with higher NVP Cmin values in individuals carrying the variant allele for either polymorphism. CYP2B6 516G>T

(adjusted p = 0.09, Figure 3.5) genotype did not meet statistical significance when corrected for the number of haplotype blocks in the gene but did show a trend toward 74 higher NVP Cmin values. Two polymorphisms in CYP2C19 (rs4917623, p = 0.03; rs4388808, p = 0.05) were also associated with higher NVP Cmin values (Figure 3.6,

Figure 3.7).

Figure 3.3. Nevirapine Cmin as a function of CYP2B6 983T>C genotype. Concentrations (mean ± SEM) are stratified by CYP2B6 983T>C genotype (adjusted p = 0.008).

75

Figure 3.4. Nevirapine Cmin as a function of ABCC10 rs2125739. Concentrations (mean ± SEM) are stratified by ABCC10 rs2125739 genotype (adjusted p = 0.009).

76

Figure 3.5. Nevirapine Cmin as a function of CYP2B6 516G>T genotype. Concentrations (mean ± SEM) are stratified by CYP2B6 516G>T genotype (adjusted p = 0.08).

77

Figure 3.6. Nevirapine Cmin as a function of CYP2C19 rs491623 genotype. Concentrations (mean ± SEM) are stratified by CYP2C19 rs491623 genotype (adjusted p = 0.03). In the TC/TT genotype group grey circles = TC, black circles = CC.

78

Figure 3.7. Nevirapine Cmin as a function of CYP2C19 rs4388808 genotype. Concentrations (mean ± SEM) are stratified by CYP2C19 rs4388808 genotype (adjusted p = 0.05). In the GA/GG genotype group black circles = GA, grey circles = GG.

79

In the multivariate analysis, CYP2B6 983T>C (p = 0.007), ABCC10 rs2125739 (p =

0.008) and age (p = 0.045) are significantly associated with NVP Cmin values (Table

3.4).

Table 3.4. Multivariate analysis of the association of NVP Cmin with Genotypes and Demographic Covariates

Independent Upper Lower 2 Beta SE P Variable1 95% CI 95% CI Age 0.002 0.002 0.006 -0.002 0.045 Gender (M) -0.03 0.04 0.049 -0.109 0.116 CYP2B6 983T>C 0.09 0.04 0.169 0.011 0.007 (CT/CC) ABCC10 rs2125739 0.09 0.03 0.149 0.031 0.008 (CT/CC) CYP2C19 rs4917623 0.05 0.06 0.169 -0.069 0.178 (CT/CC) CYP2C19 rs4388808 -0.0009 0.06 0.118 -0.120 0.998 (GA/GG) 1Beta and SE values are reported for each factor tested against the control level. Level in parentheses is the level being tested against the control level. 2Significant values are highlighted bold

3.4.4 CYP2B6 and ABCC10 Composite Genotypes have Significant Effects on

NVP Cmin:

In order to investigate the combined effects of CYP2B6 516G>T and 983T>C, composite genotypes were constructed. There was a significant effect with subjects carrying more variant alleles having higher NVP Cmin values (Figure 3.8, p = 0.0002).

Due to the significant effects seen for SNPs in CYP2B6 and ABCC10, a composite genotype that included SNPs from both genes was constructed. This CYP2B6/ABCC10

-6 composite genotype was highly correlated with NVP Cmin levels (p = 2.5x10 ; Figure

3.9).

80

Figure 3.8. Nevirapine Cmin as a function of CYP2B6 composite genotype. Nevirapine Cmin stratified by the number of variant alleles present in CYP2B6 516G>T and CYP2B6 983T>C genotypes (p = 0.0002). Dark bars indicate the mean and light bars indicate standard deviations. All subjects in the “0” group (n = 38) have GG/TT genotypes. In the “1” group (n = 53) light grey subjects are GT/TT, dark grey are GG/TC. In the “2” group (n = 27) light grey subjects are GG/CC, medium grey are TT/TT and black are GT/TC. The “3” group (n = 1) is TT/TC.

81

Figure 3.9. Nevirapine Cmin as a function of CYP2B6/ABCC10 composite genotype. Nevirapine Cmin stratified by the number of variant alleles present in CYP2B6 516G>T, CYP2B6 983T>C and ABCC10 rs2125739 genotypes (p = 2.5x10-6). Dark bars indicate the mean and light bars indicate standard deviations. 0 alleles, n = 27; 1 allele, n = 49; 2 alleles, n = 25; 3 alleles, n = 16; 4 alleles, n = 2.

3.5 Discussion:

There have been many studies that show that CYP2B6 polymorphisms are an important predictor of NNRTI pharmacokinetics. However, the majority of studies have focused on EFV, are conducted in Caucasian populations and polymorphisms are studied for their independent effects. This study explored the association of SNPs in established candidate genes and identified several for further analysis. Consistent with other studies, the importance of CYP2B6 in NVP pharmacokinetics was confirmed9,10,14,21,28. The importance of ABCC10 polymorphisms on NVP pharmacokinetics in African populations, which has only recently been demonstrated in

82

Caucasians, was also described16. Interestingly, the importance of the combined effects of polymorphisms in CYP2B6 and ABCC10 on NVP pharmacokinetics was also evident; composite phenotypes have previously only been explored for CYP2B621,29.

Soon after the approval of NVP, unexplained variation in its pharmacokinetic parameters not attributable to environmental factors was observed30. Much of this variation has been explained by CYP2B6 polymorphisms, notably CYP2B6 516G>T and

CYP2B6 983T>C5,8,10,14,29,31. This is only the second study to demonstrate the effect of a composite CYP2B6 516/983 genotype on NVP pharmacokinetics, which underlines the importance of the combined effects of both alleles for NVP pharmacokinetics21.

While other studies have investigated the effect of CYP2B6 516/983 genotypes on NVP pharmacokinetics and did not observe an effect, this may have been due to the lack of

CYP2B6 autoinduction caused by extended NVP treatment, as the populations they were studying had only received a single dose of NVP13,29.

This study is the first to demonstrate the effect of ABCC10 rs2125739 on NVP plasma concentrations in an African population. A previous study showed that NVP is a substrate for MRP7 and found an association between rs2125739 and NVP pharmacokinetics in Caucasians, but not in Africans16. It is possible that the current study was better powered to observe an effect since the sample was larger than the previous study. ABCC10 is expressed higher in the kidney than in the liver and MRP7 may efflux NVP from the kidney into the urine. In such a case, reducing function of this transporter could result in reduced renal clearance and higher plasma concentrations of

NVP32. This study confirms the importance of MRP7 in controlling NVP plasma

83 concentrations and suggests that it is worthwhile to further investigate its influence on

NVP pharmacokinetics and efficacy.

The most interesting finding in this study is the substantial combined effect of

CYP2B6/ABCC10 genotype on NVP plasma concentrations. The number of variant alleles in both genes contributed significantly to a rise in NVP plasma concentrations.

This could be due to a combination of reduced metabolism and decreased efflux of NVP into the urine. This finding could lead to more precise dosing guidelines for NVP which would increase efficacy, decrease viral resistance and decrease toxicity. The effect of multiple genotypes in several genes along with demographic parameters has successfully been employed to design the dosing regimen for other drugs, most notably warfarin33.

This study did not find any associations between NVP Cmin and CYP2C19, CYP2C9,

NR1L2, CYP2D6, CYP3A4 and CYP3A512,13,22. These genes were selected on the basis of their role in NVP pharmacokinetics. Only the CYP3A5*3 variant has previously been associated with a reduction in NVP AUC34. No effect of this allele was observed in this study, however, only NVP Cmin was evaluated.

84

3.6 Conclusions

In summary, this study found that the combined CYP2B6/ABCC10 genotype has a significant effect on NVP Cmin concentrations in Ugandan patients receiving antiretroviral treatment. Due to the high rates of HIV infection in sub-Saharan Africa and the widespread use of NVP in that region, this information could be used to tailor dosing in patients which would lead to increases in efficacy and decreases in viral resistance and toxicity.

85

3.7 References

1. UNAIDS, GLOBAL REPORT: UNAIDS report on the global AIDS epidemic 2013. (2013). 2. C. J. Hoffmann, K. L. Fielding, S. Charalambous, M. S. Sulkowski, C. Innes, C. L. Thio, R. E. Chaisson, G. J. Churchyard, A. D. Grant, Antiretroviral therapy using zidovudine, lamivudine, and efavirenz in South Africa: tolerability and clinical events. AIDS (London, England) 22, 67-74 (2008)10.1097/QAD.0b013e3282f2306e). 3. P. o. A. G. f. A. a. Adolescents, Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Department of Health and Human Services, (2013). 4. W. H. Organization. (2013). 5. M. Gandhi, L. Z. Benet, P. Bacchetti, A. Kalinowski, K. Anastos, A. R. Wolfe, M. Young, M. Cohen, H. Minkoff, S. J. Gange, R. M. Greenblatt, Nonnucleoside reverse transcriptase inhibitor pharmacokinetics in a large unselected cohort of HIV-infected women. Journal of acquired immune deficiency syndromes (1999) 50, 482-491 (2009). 6. J. van Griensven, R. Zachariah, F. Rasschaert, J. Mugabo, E. F. Atté, T. Reid, Stavudine- and nevirapine-related drug toxicity while on generic fixed-dose antiretroviral treatment: incidence, timing and risk factors in a three-year cohort in Kigali, Rwanda. Transactions of the Royal Society of Tropical Medicine and Hygiene 104, 148-153 (2010)10.1016/j.trstmh.2009.07.009). 7. J. E. Micheli, L. W. Chinn, S. B. Shugarts, A. Patel, J. N. Martin, D. R. Bangsberg, D. L. Kroetz, Measuring the overall genetic component of nevirapine pharmacokinetics and the role of selected polymorphisms: towards addressing the missing heritability in pharmacogenetic phenotypes? Pharmacogenetics and genomics, 1-6 (2013)10.1097/FPC.0b013e32836533a5). 8. A. Saitoh, E. Sarles, E. Capparelli, F. Aweeka, A. Kovacs, S. K. Burchett, A. Wiznia, S. Nachman, T. Fenton, S. A. Spector, CYP2B6 genetic variants are associated with nevirapine pharmacokinetics and clinical response in HIV-1- infected children. AIDS (London, England) 21, 2191-2199 (2007)10.1097/QAD.0b013e3282ef9695). 9. C. Wyen, H. Hendra, M. Vogel, C. Hoffmann, H. Knechten, N. H. Brockmeyer, J. R. Bogner, J. Rockstroh, S. Esser, H. Jaeger, T. Harrer, S. Mauss, J. van Lunzen, N. Skoetz, A. Jetter, C. Groneuer, G. Fätkenheuer, S. H. Khoo, D. Egan, D. J. Back, A. Owen, Impact of CYP2B6 983T>C polymorphism on non- nucleoside reverse transcriptase inhibitor plasma concentrations in HIV-infected patients. The Journal of antimicrobial chemotherapy 61, 914-918 (2008)10.1093/jac/dkn029).

86

10. S. R. Penzak, G. Kabuye, P. Mugyenyi, F. Mbamanya, V. Natarajan, R. M. Alfaro, C. Kityo, E. Formentini, H. Masur, Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients in Uganda. HIV medicine 8, 86-91 (2007)10.1111/j.1468-1293.2007.00432.x). 11. J. Bertrand, M. Chou, D. M. Richardson, C. Verstuyft, P. D. Leger, F. Mentré, A.- M. Taburet, D. W. Haas, Multiple genetic variants predict steady-state nevirapine clearance in HIV-infected Cambodians. Pharmacogenetics and genomics, 1-9 (2012)10.1097/FPC.0b013e32835a5af2). 12. P. Riska, M. Lamson, T. Macgregor, J. Sabo, S. Hattox, J. Pav, J. Keirns, DISPOSITION AND BIOTRANSFORMATION OF THE ANTIRETROVIRAL DRUG NEVIRAPINE IN HUMANS ABSTRACT :. Pharmacology 27, (1999). 13. P. Fan-Havard, Z. Liu, M. Chou, Y. Ling, A. Barrail-Tran, D. W. Haas, A.-M. Taburet, Pharmacokinetics of phase I nevirapine metabolites following a single dose and at steady state. Antimicrobial agents and chemotherapy 57, 2154-2160 (2013)10.1128/AAC.02294-12). 14. J. Chen, J. Sun, Q. Ma, Y. Yao, Z. Wang, L. Zhang, L. Li, F. Sun, H. Lu, CYP2B6 polymorphism and nonnucleoside reverse transcriptase inhibitor plasma concentrations in Chinese HIV-infected patients. Therapeutic drug monitoring 32, 573-578 (2010)10.1097/FTD.0b013e3181ea953c). 15. T. Mahungu, C. Smith, F. Turner, D. Egan, M. Youle, M. Johnson, S. Khoo, D. Back, a. Owen, Cytochrome P450 2B6 516G-->T is associated with plasma concentrations of nevirapine at both 200 mg twice daily and 400 mg once daily in an ethnically diverse population. HIV medicine 10, 310-317 (2009)10.1111/j.1468-1293.2008.00689.x). 16. N. J. Liptrott, S. Pushpakom, C. Wyen, G. Fätkenheuer, C. Hoffmann, S. Mauss, H. Knechten, N. H. Brockmeyer, E. Hopper-Borge, M. Siccardi, D. J. Back, S. H. Khoo, M. Pirmohamed, A. Owen, Association of ABCC10 polymorphisms with nevirapine plasma concentrations in the German Competence Network for HIV/AIDS. Pharmacogenetics and genomics 22, 10-19 (2012)10.1097/FPC.0b013e32834dd82e). 17. M. H. Hofmann, J. K. Blievernicht, K. Klein, T. Saussele, E. Schaeffeler, M. Schwab, U. M. Zanger, Aberrant splicing caused by single nucleotide polymorphism c.516G>T [Q172H], a marker of CYP2B6*6, is responsible for decreased expression and activity of CYP2B6 in liver. The Journal of pharmacology and experimental therapeutics 325, 284-292 (2008). 18. K. Klein, T. Lang, T. Saussele, E. Barbosa-Sicard, W.-H. Schunck, M. Eichelbaum, M. Schwab, U. M. Zanger, Genetic variability of CYP2B6 in populations of African and Asian origin: allele frequencies, novel functional variants, and possible implications for anti-HIV therapy with efavirenz. Pharmacogenetics and genomics 15, 861-873 (2005). 19. M. Arab-Alameddine, J. Di Iulio, T. Buclin, M. Rotger, R. Lubomirov, M. Cavassini, A. Fayet, L. A. Décosterd, C. B. Eap, J. Biollaz, A. Telenti, C. Csajka, Pharmacogenetics-based population pharmacokinetic analysis of efavirenz in 87

HIV-1-infected individuals. Clinical pharmacology and therapeutics 85, 485-494 (2009)10.1038/clpt.2008.271). 20. S. K. Gupta, S. L. Rosenkranz, Y. S. Cramer, S. L. Koletar, L. a. Szczech, V. Amorosa, S. D. Hall, The pharmacokinetics and pharmacogenomics of efavirenz and lopinavir/ritonavir in HIV-infected persons requiring hemodialysis. AIDS (London, England) 22, 1919-1927 (2008)10.1097/QAD.0b013e32830e011f). 21. S. G. Heil, M. E. V. D. Ende, P. W. Schenk, I. V. D. Heiden, J. Lindemans, D. Burger, R. H. N. V. Schaik, Associations Between ABCB1, CYP2A6, CYP2B6, CYP2D6, and CYP3A5 Alleles in Relation to Efavirenz and Nevirapine Pharmacokinetics in HIV-Infected Individuals. Therapeutic drug monitoring 34, 153-159 (2012). 22. M. Whirl-Carrillo, E. M. McDonagh, J. M. Hebert, L. Gong, K. Sangkuhl, C. F. Thorn, R. B. Altman, T. E. Klein, Pharmacogenomics knowledge for personalized medicine. Clin Pharmacol Ther 92, 414-417 (2012); published online EpubOct (10.1038/clpt.2012.96). 23. S. R. Faucette, T. C. Zhang, R. Moore, T. Sueyoshi, C. J. Omiecinski, E. L. LeCluyse, M. Negishi, H. Wang, Relative activation of human pregnane X receptor versus constitutive androstane receptor defines distinct classes of CYP2B6 and CYP3A4 inducers. J Pharmacol Exp Ther 320, 72-80 (2007); published online EpubJan (10.1124/jpet.106.112136). 24. H. N. Mistri, A. G. Jangid, A. Pudage, N. Gomes, M. Sanyal, P. Shrivastav, High throughput LC-MS/MS method for simultaneous quantification of lamivudine, stavudine and nevirapine in human plasma. Journal of chromatography. B, Analytical technologies in the biomedical and life sciences 853, 320-332 (2007)10.1016/j.jchromb.2007.03.047). 25. A. P. Boyle, E. L. Hong, M. Hariharan, Y. Cheng, M. A. Schaub, M. Kasowski, K. J. Karczewski, J. Park, B. C. Hitz, S. Weng, J. M. Cherry, M. Snyder, Annotation of functional variation in personal genomes using RegulomeDB. Genome Research 22, 1790-1797 (2012). 26. L. D. Ward, M. Kellis, HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Research 40, D930-934 (2011)10.1093/nar/gkr917). 27. R. C. Team, R. F. f. S. Computing, R: A Language and Environment for Statistical Computing. (Vienna, Austria, 2012). 28. A. Calcagno, A. D'Avolio, M. Simiele, J. Cusato, R. Rostagno, V. Libanore, L. Baietto, M. Siccardi, S. Bonora, G. Di Perri, Influence of CYP2B6 and ABCB1 SNPs on nevirapine plasma concentrations in Burundese HIV-positive patients using dried sample spot devices. British journal of clinical pharmacology 74, 134- 140 (2012)10.1111/j.1365-2125.2012.04163.x). 29. David W. W. Haas, T. Gebretsadik, G. Mayo, Usha N. N. Menon, Edward P. P. Acosta, A. Shintani, M. Floyd, C. M. M. Stein, Grant R. R. Wilkinson, Associations between CYP2B6 polymorphisms and pharmacokinetics after a

88

single dose of nevirapine or efavirenz in African americans. The Journal of infectious diseases 199, 872-880 (2009)10.1086/597125). 30. M. M. de Maat, A. D. Huitema, J. W. Mulder, P. L. Meenhorst, E. C. van Gorp, J. H. Beijnen, Population pharmacokinetics of nevirapine in an unselected cohort of HIV-1-infected individuals. Br J Clin Pharmacol 54, 378-385 (2002); published online EpubOct ( 31. M. Rotger, S. Colombo, H. Furrer, G. Bleiber, T. Buclin, B. L. Lee, O. Keiser, J. Biollaz, L. Décosterd, A. Telenti, Influence of CYP2B6 polymorphism on plasma and intracellular concentrations and toxicity of efavirenz and nevirapine in HIV- infected patients. Pharmacogenetics and genomics 15, 1-5 (2005). 32. M. N. McCall, K. Uppal, H. a. Jaffee, M. J. Zilliox, R. a. Irizarry, The Gene Expression Barcode: leveraging public data repositories to begin cataloging the human and murine transcriptomes. Nucleic acids research 39, D1011-1015 (2011)10.1093/nar/gkq1259). 33. E. a. Sconce, T. I. Khan, H. a. Wynne, P. Avery, L. Monkhouse, B. P. King, P. Wood, P. Kesteven, A. K. Daly, F. Kamali, The impact of CYP2C9 and VKORC1 genetic polymorphism and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen. Blood 106, 2329-2333 (2005)10.1182/blood-2005-03-1108). 34. K. C. Brown, M. C. Hosseinipour, J. M. Hoskins, R. K. Thirumaran, H.-C. Tien, R. Weigel, J. Tauzie, I. Shumba, J. K. Lamba, E. G. Schuetz, H. L. McLeod, A. D. M. Kashuba, A. H. Corbett, in Pharmacogenomics. (2012), vol. 13, pp. 113-121.

89

Chapter 4

4. Genetic Predictors of HIV-1 Induced Peripheral Neuropathy in Ugandan HIV- 1+ Subjects

4.1 Abstract

Sensory peripheral neuropathy is one of the most common neurological complications associated with HIV infection. HIV related sensory neuropathy (HIV-SN) is common in the developing world due to complications in the delivery of timely antiretroviral therapy. The objective of our study is to identify genetic predictors of HIV-

SN in a treatment naive Ugandan HIV+ population. DNA and symptom data were collected from 638 patients enrolled in a cohort study in Uganda to examine treatment outcomes in HIV+ subjects. Patients were treatment naïve at enrollment and an extensive symptom questionnaire was completed to capture HIV-SN symptoms.

Genotyping was performed using the Illumina OmniExpress® platform with 681,315

SNPs and 580 samples passing quality control. Imputation with IMPUTE2 added >16 million SNPs. HIV-SN was scored on a Likert scale and recorded prior to the initiation of antiretroviral therapy. An additive genetic model and logistic regression were used to identify genes associated with HIV-SN. Replication was performed in a new set of

Ugandan subjects for selected SNPs using Taqman genotyping assays. The genome- wide analyses identified multiple SNPs associated with HIV-induced PN, including several SNPs proximal to FOLH1 (top SNP: rs2007068, p = 2.63 x 10-7), a protein associated with folate metabolism. Replication studies and meta-analyses were performed on three SNPs that had putative biological or functional effects on FOLH1.

One SNP (rs7925419) showed a statistical trend (p = 0.12) in the replication study for

90 association with HIV-SN development. This study suggests that genetic variation in

FOLH1, a gene important to folate metabolism, may influence an individual patient’s risk of developing HIV-SN. Further studies are warranted to determine the precise role that

FOLH1 may play in the development of HIV-SN.

91

4.2 Introduction

Over 35 million people worldwide are currently living with HIV infections and 69% of these people live in sub-Saharan Africa1. The advent of Highly Active Antiretroviral

Therapy (HAART) has greatly improved patient mortality but the developing world still struggles to treat HIV patients in a timely and effective manner1,2. As a consequence, many HIV+ patients in sub-Saharan Africa experience greater morbidity and mortality than HIV+ patients in the developed world3.

HIV related sensory neuropathies (HIV-SN) are a common neurological complication of HIV infection4. Reports from the pre-HAART era indicate that up to 35% of HIV infected patients will develop HIV-SN5-7. There are several proposed mechanisms for the development of HIV-SN, the primary being macrophage invasion of the peripheral nerve and viral protein toxicity4. Although the mechanism of HIV-SN is still not fully understood, risk factors including nutritional deficiencies, alcoholism, age, disease stage, low CD4+ T-cell counts and high viral load have been extensively documented8,9.

Because antiretroviral drugs, specifically nucleoside reverse transcriptase inhibitors

(NRTIs), may also cause sensory neuropathies that are clinically indistinguishable from

HIV-SN, it is difficult to ascertain whether patients receiving HAART develop sensory neuropathies due to HIV infection or drug toxicity9. There have been several candidate gene studies to characterize the effect of specific genes on the development of peripheral neuropathy due to NRTI toxicity, however to date no genetic studies have been conducted to characterize the role of host genetics on HIV-SN10-13.

Genome-wide association studies (GWAS) have identified the role of patient genetics for many complex diseases, including HIV infection, and have emphasized the

92 importance of host genetics on HIV infection and progression14,15. However, the role of host genetics in the development of HIV-SN is still unknown.

The goal of this GWAS study was to identify genetic predictors of HIV-SN in a treatment naïve HIV+ Ugandan population. Replication and bioinformatics analyses were used to further define genomic regions likely to influence the HIV-SN phenotype.

4.3 Materials and Methods:

4.3.1 Participants

Patients were recruited from the Uganda AIDS Rural Treatment Outcomes (UARTO) and Antiretrovirals in Kaposi Sarcoma (ARKS) cohorts. Study participants were treatment naïve HIV+ patients living in Mbarra, Uganda and Kampala, Uganda, respectively. Patients enrolled in the study received a treatment regimen consisting of two nucleoside reverse transcriptase inhibitors (NRTI) and one non-nucleoside reverse transcriptase inhibitor (NNRTI) or one protease inhibitor (PI). Whole blood or saliva samples were obtained at enrollment and shipped to the University of California San

Francisco for DNA isolation. Study visits were conducted upon enrollment and every three months thereafter and consist of an extensive symptom interview, CD4+ T-cell counts and viral load measurements. Genotype and phenotype data were collected on

638 patients for the initial study and 209 patients for the replication study.

93

4.3.2 Genotyping

In the discovery study, genomic DNA from 638 subjects was extracted from either saliva or whole blood samples using standard DNA extraction techniques. DNA concentrations were determined by Quant-iT™ PicoGreen® dsDNA Assays (Life

Technologies, Grand Island, NY) performed according to the manufacturer’s instructions and samples were normalized in Tris-EDTA buffer to a concentration of 50 ng/µl and stored at -80°C. Genotyping was performed on the HumanOmniExpress BeadChip

(Illumina, San Diego, CA), which interrogated 733,202 SNPs. Subjects with genotyping call rates < 95% (n = 19) were excluded from further analysis and the remaining genotypes were reclustered using BeadStudio data analysis software

(Illumina, San Diego, CA). Gender was determined from the genotype data using

PLINK software (http://pngu.mgh.harvard.edu/purcell/plink/) and compared to the database recorded sex to ensure concordance16. Eleven samples did not have concordant genders and were excluded from further analysis. As closely related individuals may confound downstream statistical analyses, Identify by Descent (IBD) was determined using PLINK 16. Twenty-five individuals were found to be greater than

12.5% related and were investigated to determine the source of the relatedness. In the instances where it was possible to establish legitimate relatedness, the subject with better quality genotyping was retained for further analysis, otherwise the samples were excluded from further analysis. In total, 15 subjects were excluded due to relatedness issues. To further ensure sample quality, genomic heterozygosity was evaluated using

PLINK and two subjects were excluded due to heterozygosity issues16. Six subjects were excluded from downstream analyses due to study related issues such as

94 withdrawal from the study. A total of 585 subjects passed genotyping and subject quality control procedures. A summary of quality control procedures for samples and single nucleotide polymorphisms (SNPs) can be viewed in Table 4.1.

Table 4.1. Quality Control of Genotype and Subject Data

Parameter Potential Cause # Excluded # Remaining

Pre QC Dataset - 0 645 Genotyping Call Rate Poor quality DNA 19 626 (>95%) Sex Concordance Database error/sample mix-up 11 615 Relatedness (IBD) Subjects are related/Poor quality DNA 15 600 (<12.5% related) Heterozygosity Poor quality DNA 2 598 Defines ethnicity/Used to correct for Population Stratification 0 598 differences in ethnicities e.g., Subject withdrew from study, Study Related Issues 6 592 inadequate longitudinal data Genotyping HAPMAP trios included to verify Controls/Duplicate 7 585 accuracy and precision of genotyping Sample Final Dataset - 585

To ensure SNP quality, filters for SNPs with call rates of greater than 95% and greater than 1% minor allele frequency were applied using PLINK software16. Out of the

733,202 SNPs genotyped, 681,315 passed quality control procedures (Table 4.2).

95

Table 4.2. Quality Control of Genotype Data

Parameter # SNPs Excluded # SNPs Remaining Pre-QC 0 733,202 Call rate > 95% 12,356 720,846 MAF ≥ 1% 39,531 681,315

To ensure that population stratification due to ethnicity did not introduce bias into downstream analyses, a principal components analysis comparing the study cohort with

HAPMAP world populations was performed using Eigenstrat software17,18. All subjects clustered near the HAPMAP African population samples (Figure 4.1).

96

Figure 4.1. Principal component analysis of study samples compared to world HAPMAP populations. Red circles are Utah residents (CEPH) with Northern and Western European ancestry (CEU), pink circles are Han Chinese in Bejing, China (CHB), blue circles are Gujarati Indian from Houston, Texas (GIH), purple circles are Yoruba in Ibadan, Nigera (YRI), black circles are ARKS and green are UARTO populations.

97

Genotyping for the replication study was performed according to the manufacturer’s instructions using the SNP specific Taqman Genotyping assays (Life Technologies,

Grand Island, NY) outlined in Table 4.3.

Table 4.3. SNPs Selected for Replication

SNP Assay ID Tag SNPs (r2 >0.8) rs7925419 Custom: AHLJXQZ rs2007068 Custom: AHI12Y6 rs9332434, rs7937386, rs2007090 rs11245616 Custom: AHKAZKR rs11245609, rs12361625 Taqman assay IDs are listed. SNPs that are in LD (r2 >0.8) in the LWK 1000 Genomes database are also listed.

Imputation was performed using SHAPEIT (http://www.shapeit.fr/) and IMPUTE2

(http://mathgen.stats.ox.ac.uk/impute/impute_v2.html) software. Prephasing of haplotypes was performed according to the developer’s instructions for prephasing for imputation using SHAPEIT version 119. Imputation was performed on samples and

SNPs passing quality control steps in IMPUTE2 using as a reference panel the 1000

Genomes Phase I integrated variant set which includes 1,092 individuals from Africa,

Asia, Europe, and the Americas20. Approximately 40 million SNPs were imputed which were subsequently filtered for quality (info score > 0.8), leaving 16.9 million SNPs for analysis21. Because this dataset is intended for further use in additional studies, filtering

SNPs for MAF > 1% was performed after statistical analyses were performed.

4.3.3 Phenotype

Peripheral neuropathy data were gathered during the study visits as a component of the symptom questionnaire. Peripheral neuropathy is graded on a Likert-type scale with subjects asked if they are experiencing “pain, numbness or tingling in the hands or feet.”

A score of “0” denotes no symptoms, “1” means “bothers me not at all”, “2” means

98

“bothers me a little”, “3” denotes “bothers me a moderate amount” and “4” means

“bothers me a lot”. Subjects who reported no symptoms were assigned as “controls” (n

= 342). Subjects who reported symptoms in the 3 and 4 categories were assigned as

“cases” (n = 129).

4.3.4 Statistical Analyses

Statistical analyses to determine the effect of demographic covariates were performed in R using ANOVA for continuous variables and Chi-squared tests for categorical variables22. Standard case/control analyses using logistic regression were performed for the primary and replication analyses using PLINK to test the association between each SNP and the phenotype16. Odds ratios (OR), 95% confidence intervals

(95% CI) and p-values were generated for each SNP. To account for the multiple testing burden genome-wide significance was considered p ≤ 5 x 10-8 and genome-wide suggestive was considered p ≤ 5 x 10-7; ‘promising’ SNPs at p ≤ 5 x 10-6 were considered for further bioinformatic analysis23. Allele frequencies in the control group for

SNPs of interest were examined to ensure they were in Hardy-Weinberg Equilibrium using a Chi-squared test. Linkage disequilibrium calculations were performed in

PLINK16. Plots were produced in R and Microsoft Excel 201022. Meta analyses to combine p-values from the discovery and replication studies were performed in R using the meta package24.

4.3.5 Bioinformatic Analyses

To explore the putative biological significance of SNPs that had ‘promising’ p-values

(p ≤ 5 x 10-6), these SNPs were selected for further bioinformatic analysis. SNPs were

99 annotated to genes using SNPnexus and the UCSC genome browser25,26. GTEx

(http://www.broadinstitute.org/gtex/) and GeneVar

(http://www.sanger.ac.uk/resources/software/genevar/) databases were employed to examine the effect of a SNP on gene expression27. To determine regulatory functions of SNPs of interest, Haploreg and ENCODE databases were employed28,29. SNPs with

‘promising’ p-values that also had putative biological function or regulatory consequences were chosen for replication.

4.4 Results

4.4.1 Demographic Data

The demographic characteristics of the discovery and replication cohorts are described in Table 4.4. The two cohorts were similar in all demographic variables assessed. None of the demographic variables tested had an effect on case/control status with the exception of gender in the discovery cohort (p < 0.001). However, this effect was not seen in the replication cohort (p = 0.8). The case percentages in the initial and replication cohorts were 27% and 23%, respectively.

Table 4.4. Patient Demographic Data in the Discovery and Replication Cohorts

Discovery Replication

Sample size, n 471 157 Age, years, median ± SD 34 ± 8.53 33 ± 9.25 Gender, n (%M) 177 (38%) 66 (42%) Baseline CD4+ T-cell counts, cells/mm3, median ± SD 133 ± 117 200 ± 204 Baseline Viral Load, log(copies/mL), median ± SD 5.20 ± 0.68 5.07 ± 0.67 Case, n (%) 129 (27%) 36 (23%) Cohort, n (%ARKS) 139 (30%) 45 (35%)

100

4.4.2 Loci in Chromosome 11 are associated with HIV induced peripheral

neuropathy

After quality control procedures, 16.9 million imputed and genotyped SNPs were tested for association with HIV-SN using a standard case/control testing methodology.

Figure depicts the p-value for each SNP plotted at their position in the genome. The Q-

Q plot and genomic inflation factor, λ (1.04, Figure 4.3), indicate that there is no significant population stratification.

101

102

Figure 4.2. Manhattan plot showing the distribution along the human autosomes of -log10 (p values) obtained for SNP association with HIV-SN cases versus control subjects. The lower red line (p < 5 x 10-6) marks the cutoff for ‘promising’ SNPs considered for further bioinformatics analysis. The upper red line marks the genome-wide suggestive cutoff of p < 5 x 10-7.

102

Figure 4.3

Figure 4.3. Observed versus expected p-values (-log base 10 scale) for SNP association with HIV-SN cases versus control subjects. Red line indicates the null distribution. The genomic inflation factor λ was 1.04 and was calculated using R statistical computing software.

Only 142 SNPs had p-values < 5 x 10-6. SNPs with the greatest association to HIV-

SN were filtered for linkage disequilibruim (r2 > 0.8) with the SNP with the lowest p- value for each LD block being retained (Bold SNPs, Table 4.5). The SNP with the most significant p-value was on chromosome 11 (rs2007068, p = 2.63 x 10-7, OR =

2.69, 95% CI 3.90 - 1.85). A robust peak can be seen at chromosome 11 with sporadic signals from other regions of the genome.

103

Table 4.5. Top Variants Associated with HIV Induced Peripheral Neuropathy

Chr SNP Frq OR U95CI L95CI P Gene Feature Left Gene Right Gene 11 rs2007068 0.15 2.69 3.90 1.85 2.63x10-7 - Intergenic OR4C12 OR4A5 11 rs12288743 0.23 2.41 3.43 1.69 5.88x10-7 - Intergenic OR4C12 OR4A5 11 rs8189012 0.43 2.20 3.01 1.61 7.21x10-7 - Intergenic OR4C12 OR4A5 11 rs2512730 0.30 2.13 2.86 1.59 7.24x10-7 OR5W2 5’ upstream OR5W2 OR5I1 11 rs146970082 0.28 2.29 3.20 1.64 7.62x10-7 - Intergenic OR4C12 OR4A5 11 rs11246460 0.27 2.29 3.20 1.64 7.72x10-7 - Intergenic OR4C12 OR4A5 11 rs4939007 0.33 2.15 2.94 1.57 8.42x10-7 - Intergenic OR5D16 TRIM51 19 rs10403857 0.33 0.39 0.57 0.27 9.24x10-7 OR1I1 5' upstream CASP14 OR1I1 11 rs2457232 0.30 2.11 2.83 1.57 9.40x10-7 - Intergenic OR5W2 OR5I1 11 rs10895994 0.19 2.41 3.43 1.69 9.53x10-7 - Intergenic OR5BE1P OR8I2 11 rs4939008 0.30 2.11 2.83 1.57 1.01x10-6 - Intergenic OR5W2 OR5I1 11 rs2449144 0.19 2.36 3.36 1.66 1.11x10-6 - Intergenic OR5F1 OR5F2P 11 rs4261267 0.37 2.10 2.82 1.57 1.14x10-6 - Intergenic OR4C12 OR4A5 11 rs4963111 0.11 3.06 4.80 1.95 1.19x10-6 - Intergenic OR4C12 OR4A5 11 rs11512987 0.11 3.05 4.79 1.94 1.20x10-6 - Intergenic OR4C12 OR4A5 14 rs75721606 0.16 2.70 4.07 1.79 1.27x10-6 TC2N Intronic CATSPERB FBLN5 11 rs7294146 0.11 3.05 4.79 1.94 1.28x10-6 - Intergenic OR4C12 OR4A5 11 rs2512928 0.20 2.31 3.22 1.66 1.39x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs2512925 0.20 2.30 3.21 1.65 1.42x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs4598652 0.38 2.07 2.78 1.54 1.47x10-6 - Intergenic OR4C12 OR4A5 11 rs4423169 0.38 2.07 2.78 1.54 1.47x10-6 - Intergenic OR4C12 OR4A5 11 rs4268491 0.38 2.07 2.78 1.54 1.47x10-6 - Intergenic OR4C12 OR4A5 11 rs2512759 0.20 2.30 3.21 1.65 1.47x10-6 - Intergenic OR5AS1 OR5AQ1P 11 11-55777327 0.20 2.33 3.32 1.64 1.47x10-6 - Intergenic OR5F1 OR5S1 11 rs6485948 0.25 2.32 3.24 1.66 1.48x10-6 - Intergenic OR4A47 TRIM49B 11 rs4363603 0.25 2.28 3.18 1.63 1.50x10-6 - Intergenic OR4A47 TRIM49B 11 rs7118155 0.20 2.34 3.33 1.64 1.50x10-6 - Intergenic OR5J1P OR8I2 11 rs148900318 0.22 2.51 3.64 1.73 1.59x10-6 - Intergenic OR4C12 OR4A5 11 rs8189038 0.38 2.05 2.75 1.53 1.61x10-6 - Intergenic OR4C12 OR4A5 11 rs2460196 0.20 2.30 3.21 1.65 1.62x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs4451712 0.38 2.05 2.75 1.53 1.63x10-6 - Intergenic OR4C12 OR4A5 11 rs4256959 0.38 2.05 2.75 1.53 1.63x10-6 - Intergenic OR4C12 OR4A5 11 rs2512966 0.20 2.30 3.21 1.65 1.63x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs8189236 0.37 2.04 2.74 1.52 1.66x10-6 - Intergenic OR4C12 OR4A5 11 rs11245616 0.12 2.92 4.49 1.90 1.67x10-6 - Intergenic OR4C12 OR4A5 11 rs2512740 0.20 2.29 3.20 1.64 1.67x10-6 - Intergenic OR5F2P OR5AS1 11 rs11518847 0.12 2.84 4.37 1.85 1.71x10-6 - Intergenic OR4C12 OR4A5 4 rs13148227 0.05 5.45 11.04 2.69 1.78x10-6 - Intergenic SMIM20 RBPJ 2 rs4848126 0.31 2.35 3.34 1.65 1.81x10-6 GLI2 Intronic FLJ14816 TFCP2L1

104

Chr SNP Frq OR U95CI L95CI P Gene Feature Left Gene Right Gene 11 rs7925419 0.14 2.67 4.03 1.77 1.82x10-6 FOLH1 5' upstream FOLH1 OR4C13 11 rs10902288 0.37 2.07 2.78 1.54 1.84x10-6 - Intergenic OR4C12 OR4A5 11 rs2851533 0.26 2.23 3.11 1.60 1.84x10-6 - Intergenic TRIM64C FOLH1 11 rs1600823 0.20 2.30 3.21 1.65 1.89x10-6 - Intergenic OR5F1 OR5F2P 11 rs4080494 0.12 2.79 4.29 1.81 1.91x10-6 - Intergenic OR4C12 OR4A5 11 rs11246421 0.37 2.05 2.75 1.53 1.91x10-6 - Intergenic OR4C12 OR4A5 11 rs8189029 0.37 2.04 2.74 1.52 1.92x10-6 - Intergenic OR4C12 OR4A5 11 rs4515955 0.37 2.04 2.74 1.52 1.92x10-6 - Intergenic OR4C12 OR4A5 11 rs4350349 0.37 2.04 2.74 1.52 1.92x10-6 - Intergenic OR4C12 OR4A5 11 rs2449126 0.20 2.30 3.27 1.62 1.92x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs12361625 0.12 2.87 4.42 1.86 1.94x10-6 - Intergenic OR4C12 OR4A5 11 rs11245609 0.12 2.87 4.42 1.86 1.94x10-6 - Intergenic OR4C12 OR4A5 11 rs7952311 0.28 2.12 2.90 1.55 1.96x10-6 - Intergenic OR5I1 OR10AF1P 11 rs12360596 0.37 2.04 2.74 1.52 1.99x10-6 - Intergenic OR4C12 OR4A5 11 rs10902283 0.37 2.04 2.74 1.52 1.99x10-6 - Intergenic OR4C12 OR4A5 11 rs4881692 0.12 2.87 4.42 1.86 2.02x10-6 - Intergenic OR4C12 OR4A5

-6 3' 11 rs2512941 0.20 2.25 3.14 1.61 2.04x10 OR5F1 OR7E5P OR5F1 downstream 3' 11 rs2460207 0.20 2.29 3.20 1.64 2.09x10-6 OR5F1 OR7E5P OR5F1 downstream 11 rs2512942 0.20 2.27 3.17 1.63 2.10x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs2460195 0.20 2.28 3.18 1.63 2.10x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs1164685 0.20 2.41 3.50 1.66 2.12x10-6 - Intergenic FOLH1 OR4C13 11 rs2449124 0.20 2.27 3.17 1.63 2.13x10-6 - Intergenic OR5AS1 OR5AQ1P 11 rs6591816 0.20 2.28 3.18 1.63 2.14x10-6 - Intergenic OR10AG1 OR7E5P 11 rs6591812 0.20 2.28 3.18 1.63 2.14x10-6 - Intergenic OR10AG1 OR7E5P 11 11-50652459 0.30 2.18 2.98 1.59 2.14x10-6 - Intergenic OR4C12 OR4A5 11 rs7939844 0.28 2.11 2.89 1.54 2.17x10-6 - Intergenic OR5I1 OR10AF1P 11 rs4103567 0.13 2.66 4.01 1.76 2.40x10-6 - Intergenic FOLH1 OR4C13 11 rs2460204 0.20 2.27 3.17 1.63 2.43x10-6 - Intergenic OR7E5P OR5F1 11 rs7294221 0.42 2.09 2.86 1.53 2.44x10-6 - Intergenic OR4C12 OR4A5 11 rs10902013 0.07 4.22 7.75 2.30 2.46x10-6 - Intergenic OR4C12 OR4A5 11 rs7927383 0.34 2.02 2.71 1.51 2.58x10-6 - Intergenic OR5I1 OR10AF1P

-6 3' 11 rs10895998 0.20 2.26 3.15 1.62 2.59x10 OR8I2 OR8I2 OR8I4P downstream 19 rs7249208 0.33 0.41 0.59 0.28 2.67x10-6 - Intergenic CASP14 OR1I1 6 rs686070 0.11 2.84 4.37 1.85 2.69x10-6 - Intergenic OFCC1 TFAP2A 11 rs59591431 0.30 2.05 2.75 1.53 2.69x10-6 - Intergenic OR5D16 OR9M1P

-6 Synonymou 11 rs11231253 0.30 2.05 2.75 1.53 2.72x10 OR5D16 OR5L2 OR9M1P s 5 rs11134288 0.47 2.16 2.96 1.58 2.80x10-6 - Intergenic MTRR SEMA5A 11 rs1164666 0.20 2.37 3.37 1.67 2.87x10-6 - Intergenic FOLH1 OR4C13 11 rs8188994 0.33 2.08 2.85 1.52 2.88x10-6 - Intergenic OR4C12 OR4A5

-6 PPP1R1 19 rs6966 0.34 2.08 2.85 1.52 2.97x10 3' UTR ERCC2 CD3EAP 3L 11 rs8189086 0.44 2.13 2.91 1.56 2.97x10-6 - Intergenic OR4C12 OR4A5

105

Chr SNP Frq OR U95CI L95CI P Gene Feature Left Gene Right Gene 11 rs10839243 0.30 2.31 3.29 1.62 3.01x10-6 - Intergenic FOLH1 OR4C13 11 rs4556555 0.20 2.29 3.26 1.61 3.11x10-6 - Intergenic OR5D16 TRIM51 11 rs7294283 0.40 2.02 2.71 1.51 3.16x10-6 - Intergenic OR4C12 OR4A5 11 rs76392625 0.40 2.03 2.72 1.51 3.21x10-6 - Intergenic OR4C12 OR4A5 11 rs10902312 0.45 2.10 2.87 1.53 3.30x10-6 - Intergenic OR4C12 OR4A5 11 rs10839211 0.21 2.41 3.50 1.66 3.31x10-6 - Intergenic OR4A47 TRIM49B 11 rs7942630 0.21 2.35 3.34 1.65 3.32x10-6 - Intergenic FOLH1 OR4C13 11 11-50207195 0.23 2.26 3.22 1.59 3.41x10-6 - Intergenic OR4C12 OR4A5 6 rs74609646 0.14 2.80 4.31 1.82 3.43x10-6 PDE7B Intronic GAPDHL19 MTFR2 19 rs7245995 0.22 0.32 0.52 0.20 3.50x10-6 - Intergenic CASP14 OR1I1 11 rs2512734 0.29 2.03 2.72 1.51 3.50x10-6 - Intergenic TRIM51 OR5W2 11 rs28582835 0.07 4.07 7.33 2.26 3.58x10-6 - Intergenic OR4C12 OR4A5 11 rs28456013 0.07 4.04 7.27 2.24 3.60x10-6 - Intergenic OR4C12 OR4A5 11 rs950413 0.07 4.00 7.20 2.22 3.64x10-6 - Intergenic OR4C12 OR4A5 11 rs76892297 0.29 2.11 2.89 1.54 3.68x10-6 - Intergenic TRIM51 OR5W2 11 11-50760329 0.46 2.11 2.89 1.54 3.69x10-6 - Intergenic OR4C12 OR4A5 11 11-51241844 0.39 2.00 2.68 1.49 3.70x10-6 - Intergenic OR4C12 OR4A5 11 rs11827319 0.40 2.01 2.70 1.50 3.73x10-6 - Intergenic OR4C12 OR4A5 11 rs8188846 0.33 2.05 2.81 1.50 3.76x10-6 - Intergenic OR4C12 OR4A5 11 11-49034349 0.21 2.40 3.48 1.65 3.78x10-6 - Intergenic OR4A47 TRIM49B 11 rs10901992 0.07 3.89 6.87 2.20 3.78x10-6 - Intergenic OR4C12 OR4A5 11 rs2193303 0.33 2.07 2.83 1.51 3.85x10-6 - Intergenic OR4C12 OR4A5 11 rs115617184 0.40 2.01 2.70 1.50 3.89x10-6 - Intergenic OR4C12 OR4A5 11 rs56080733 0.21 2.40 3.48 1.65 3.98x10-6 - Intergenic OR4A47 TRIM49B 11 rs8189163 0.40 2.01 2.70 1.50 4.00x10-6 - Intergenic OR4C12 OR4A5 18 rs1701503 0.13 2.62 3.95 1.74 4.01x10-6 - Intergenic DLGAP1 PPIAP14 11 rs7294222 0.40 2.01 2.70 1.50 4.04x10-6 - Intergenic OR4C12 OR4A5 11 rs12274709 0.40 2.01 2.70 1.50 4.04x10-6 - Intergenic OR4C12 OR4A5 11 rs7294240 0.40 2.01 2.70 1.50 4.07x10-6 - Intergenic OR4C12 OR4A5 11 rs7294226 0.40 2.00 2.68 1.49 4.08x10-6 - Intergenic OR4C12 OR4A5 11 rs142781985 0.43 2.07 2.83 1.51 4.08x10-6 - Intergenic OR4C12 OR4A5 6 rs188640115 0.14 2.70 4.16 1.75 4.10x10-6 PDE7B Intronic GAPDHL19 MTFR2 11 rs2851564 0.26 2.20 3.07 1.58 4.12x10-6 - Intergenic TRIM64C FOLH1 11 rs2727015 0.26 2.20 3.07 1.58 4.12x10-6 - Intergenic TRIM64C FOLH1 11 rs12224686 0.29 2.02 2.71 1.51 4.16x10-6 TRIM51 Intronic OR5D6 OR5W1P 11 rs4553353 0.40 2.00 2.68 1.49 4.19x10-6 - Intergenic OR4C12 OR4A5 11 rs4393301 0.40 2.00 2.68 1.49 4.20x10-6 - Intergenic OR4C12 OR4A5 11 rs8189043 0.40 2.01 2.70 1.50 4.21x10-6 - Intergenic OR4C12 OR4A5 6 rs9376172 0.14 2.69 4.14 1.75 4.23x10-6 PDE7B Intronic GAPDHL19 MTFR2 11 rs4362131 0.40 2.01 2.70 1.50 4.24x10-6 - Intergenic OR4C12 OR4A5 11 rs7294255 0.40 2.01 2.70 1.50 4.26x10-6 - Intergenic OR4C12 OR4A5 11 rs58036605 0.07 3.86 6.81 2.19 4.30x10-6 - Intergenic OR4C12 OR4A5

106

Chr SNP Frq OR U95CI L95CI P Gene Feature Left Gene Right Gene 11 rs11245788 0.07 3.86 6.81 2.19 4.30x10-6 - Intergenic OR4C12 OR4A5 11 rs8188898 0.35 2.08 2.85 1.52 4.34x10-6 - Intergenic OR4C12 OR4A5

-6 TMPRSS11 4 rs28633232 0.23 2.21 3.08 1.58 4.35x10 - Intergenic YTHDC1 B 6 rs672962 0.14 2.60 3.92 1.72 4.36x10-6 - Intergenic OFCC1 TFAP2A 4 rs11940780 0.23 2.21 3.08 1.58 4.37x10-6 - Intergenic TMPRSS11B YTHDC1 11 rs2512736 0.20 2.23 3.17 1.57 4.47x10-6 - Intergenic OR5W1P OR5W2 6 rs4896193 0.14 2.68 4.04 1.78 4.50x10-6 PDE7B Intronic GAPDHL19 MTFR2 11 rs1396625 0.20 2.23 3.17 1.57 4.50x10-6 - Intergenic OR5W1P OR5W2 11 11-50464664 0.26 2.16 3.01 1.55 4.51x10-6 - Intergenic OR4C12 OR4A5 11 rs17492683 0.19 2.35 3.41 1.62 4.62x10-6 - Intergenic OR4A16 OR4A15 11 11-48945608 0.30 2.24 3.19 1.57 4.68x10-6 - Intergenic OR4A47 TRIM49B 3 rs6779831 0.35 0.45 0.64 0.32 4.72x10-6 - Intergenic GRM7 LMCD1 11 rs61897488 0.20 2.27 3.23 1.60 4.72x10-6 - Intergenic OR5D16 TRIM51 11 rs61897487 0.20 2.27 3.23 1.60 4.72x10-6 - Intergenic OR5D16 TRIM51 11 11-55686442 0.32 2.08 2.85 1.52 4.73x10-6 - Intergenic OR5W2 OR5I1 11 rs8188856 0.45 2.06 2.82 1.51 4.81x10-6 - Intergenic OR4C12 OR4A5 11 rs11245888 0.42 2.03 2.72 1.51 4.81x10-6 - Intergenic OR4C12 OR4A5 11 rs10444243 0.07 3.96 7.13 2.20 4.82x10-6 - Intergenic OR4C13 OR4C12 4 4-25982104 0.11 3.00 4.80 1.87 4.92x10-6 - Intergenic SMIM20 RBPJ 11 rs2512726 0.20 2.21 3.08 1.58 4.99x10-6 - Intergenic OR5W2 OR5I1 The lowest p-value SNP for each LD block is highlighted in bold. Upper 95% Confidence interval, U95CI; Lower 95% Confidence interval, L95CI.

4.4.3 Polymorphisms on chromosome 11 may have an effect on FOLH1

regulation and expression

The region around the most significant SNPs associated with HIV-SN on chromosome 11 has a multitude of genes, with the majority being olfactory receptors and uncharacterized genes (Error! Reference source not found. 4.4). Of biological interest in the region was folate hydrolase 1 (FOLH1), which catalyzes the step to transform dietary folate to folic acid prior to the entry of folic acid into the folate cycle.

All of the SNPs on chromosome 11 that had p-values < 5 x 10-6 clustered within four independent loci near FOLH1 (Figure 4.5). Three of the loci were at the 5’ end of the gene and one was at the 3’ end of the gene. Initially the SNP closest to FOLH1

107

(rs7925419, 1.6 kb from the 5’-end of FOLH1) was analyzed for putative functional or regulatory effects.

108

109

Figure 4.4. UCSC genome browser image showing position of SNPs associated with HIV-SN susceptibility in the vicinity of the FOLH1 gene. SNPs in the FOLH1 region track were implicated with HIV-SN case vs. control status (p < 1 x 10-5). SNPs in purple (p < 5 x 10-6) were considered ‘promising’ and selected for replication. Rs12288743 is further upstream and is not visible in this figure.

109

Figure 4.5. Loci proximal to FOLH1 associated with HIV-SN. UCSC browser detail of SNPs associated with HIN-SN (p < 10-5) proximal to FOLH1. The four independent loci are outlined in red.

The FOLH1 SNP rs7925419 had a minor allele frequency of 11% and 23% in control and case groups, respectively (Figure 4.6). The control group minor allele frequency of

11% was comparable to the 1000 Genomes project’s African population minor allele frequency for this SNP of 12%30.

110

0.35

0.30

0.25

0.23

0.20

0.15

Minor Allele MinorAllele Frequency 0.11 0.10

0.05

0.00 Control Cases Case/Control Status

Figure 4.6. Plot of the association of FOLH1 SNP rs7925419 with HIV-SN case and control status in HIV+ Ugandan subjects. Minor allele frequencies of rs7925419 in the discovery cohort are significantly higher (p = 1.82 x 10-6) in case subjects (n=129, 23%) versus control subjects (n=342, 11%).

111

Analysis for regulatory function using the Haploreg database indicated that rs7925419 is a weak enhancer in several neuronal tissues (brain inferior temporal lobe, cingulate gyrus, hippocampus and substantia nigra) and alters the binding affinity for the transcription factor NF-I28. The association between rs7925419 and FOLH1 RNA expression was explored in the GTEx database and the variant allele was associated with decreased FOLH1 RNA expression in brain cervical cord C-1 tissue (Figure 4.7, p

= 0.04).

Figure 4.7. The effect of rs7925419 genotype on RNA expression of FOLH1 in brain spinal cord cervical c1 tissue. FOLH1 levels decrease significantly with the number of variant alleles present in the genotype (p < 0.04). Data are from RNA-seq (log[RPKM]; n = 17) and was provided by the GTEx bioinformatic database. RPKM = reads per kilobase per million.

112

Additional analyses were conducted on the lowest p-value SNPs from each of the remaining three loci proximal to FOLH1. The top SNP associated with HIV-SN is rs2007068 which is 865 kb from the 5’ region of FOLH1. Regulatory functional analyses performed in Haploreg indicate that this SNP is an expression quantitative loci (eQTL) in the Gibbs Frontal cortex28. GTEx data indicates that the variant allele is associated with a reduction in RNA expression of FOLH1 in brain cervical cord C-1 tissue (p = 3 x 10-4).

The second most associated SNP with HIV-SN was rs12288743, which is in the loci most distal to FOLH1, being 1180 kb from the 5’ end of FOLH1. While regulatory analyses in the Haploreg database did not reveal any effects due to the SNP, GTEx expression analyses found that the variant allele is significantly associated with FOLH1 expression in the brain cervical cord C-1 tissue (p = 3 x 10-4).

4.4.4 Replication Results

An additional 157 subjects from the original UARTO and ARKS cohorts were available for the replication study. Three SNPs that had ‘promising’ p-values and also showed the highest potential for regulatory and functional effects were chosen for replication and genotyped using Taqman genotyping assays. A meta-analysis was also performed to see the overall effect of these SNPs in both the discovery and replication cohorts. The results of the replication study and the meta-analysis are shown in Table

4.6. Of these SNPs, rs7925419 showed a trend toward significance in the replication study (p = 0.12) and had an improved p-value in the meta-analysis (7.60 x 10-7). The two other SNPs chosen for replication showed no trend towards significance and did not have improved p-values in the meta-analysis.

113

Table 4.6. Results from replication and meta-analyses

Discovery Replication Meta-analysis Minor U95 L95 U95% L95 U95 L95 SNP Chr Gene OR P OR P OR P* Allele %CI %CI CI %CI %CI %CI

rs7925419 11 FOLH1 T 2.67 4.03 1.77 1.82x10-6 1.74 3.52 0.86 0.12 2.39 3.42 1.68 7.60x10-7

rs20087068 11 T 2.69 3.90 1.85 2.63x10-7 1.11 2.31 0.56 0.77 2.23 3.10 1.61 9.21x10-7

rs11245616 11 A 2.92 4.49 1.90 1.67x10-6 0.72 2.21 0.24 0.79 2.40 3.65 1.61 1.07x10-5

* Meta-analysis p-values reflect the one-sided p-value because the effect direction was assumed a priori to be the same direction in the replication cohort as the discovery cohort.

114

114

4.5 Discussion

HIV related sensory neuropathies are a common complication to HIV infection, particularly in the developing world3,4. Because of the rarity of HIV-SN in the developed world, little research has been conducted into the role host genetics play in its development. This study used a genome-wide association study to determine the genetic predictors of HIV-SN in a Ugandan HIV+ treatment naïve population. While no

SNPs reached genome-wide significance, multiple loci of suggestive or ‘promising’ significance were identified on chromosome 11 proximal to the FOLH1 gene.

Bioinformatic analyses indicated that three SNPs had the potential to modulate FOLH1 expression or function and these SNPs were chosen for independent replication in a subset of patients from the same cohorts as the discovery study. Replication identified one SNP (rs7925419) that showed a trend towards significance (p = 0.12) and meta- analyses of the discovery and replication cohorts showed improved statistical significance (7.60 x 10-7). The lack of statistical significance may be due to the small sample sizes of the discovery and replication cohorts; however the results for rs7925419 are intriguing. While this SNP is upstream from the 5’ end of FOLH1 it has been shown in vitro to be in the promoter region of FOLH1 and therefore may affect

FOLH1 expression as supported by data obtained from the GTEx database31.

FOLH1 is a folate hydrolase that converts dietary folate to folic acid before folic acid enters the one-carbon cycle. Polymorphisms in FOLH1 have been shown to influence plasma levels of folate and have been implicated in neural tube defects in developing fetuses32,33. Sensory peripheral neuropathies may be caused by many diseases and drug toxicities and folate-responsive peripheral neuropathies have been extensively

115 documented in the scientific literature34-37. Generally, these folate-responsive neuropathies are due to dietary deficiency or impaired folate metabolism, as is seen in peripheral neuropathy in alcoholic patients36. It has also been reported that HIV patients commonly are folate deficient38,39.

Several lines of evidence support a potential role for folate in HIV-induced SN. First, folate levels have been found to impact the development of sensory neuropathies in other diseases and HIV infection may lead to folate deficiency35-39. In addition, FOLH1 polymorphisms have been shown to effect folate plasma levels33,40,41. The rs7925419

FOLH1 SNP identified in this study has been associated with decreased expression of

FOLH1 in brain tissue and has the potential to reduce the amount of folate that is converted to dietary folate. This is the same mechanism of folate deficiency that is observed in some alcoholism related neuropathies where the absorption of folate is impaired prior to entry into the one-carbon metabolic cycle42,43.

The primary limitation of this study is the small sample sizes of the discovery and replication cohorts which decreases the power to detect variants with small effect sizes.

However, this study does provide an interesting potential mechanism of HIV-SN that could be verified with additional clinical and functional studies. Another potential limitation of this study is the use of a symptom questionnaire in lieu of a clinical test to define the HIV-SN phenotype. It has been previously demonstrated though that single question neuropathy screens (SQNS) are remarkably specific (80.0%) and sensitive

(95.7%) for the diagnosis of HIV related neuropathies44.

116

4.6 Conclusion

In conclusion, this study identified a potential genetic predictor for the development of HIV-SN, FOLH1. Since the exact role that FOLH1 and FOLH1 polymorphisms play in the development of HIV-SN is not known, further in vitro and clinical studies are warranted. These findings may lead to a potentially simple and cost effective treatment for HIV-SN, folate supplementation in HIV+ patients.

117

4.7 References

1. UNAIDS, GLOBAL REPORT: UNAIDS report on the global AIDS epidemic 2013. (2013). 2. M. Callaghan, N. Ford, H. Schneider, Review A systematic review of task-shifting for HIV treatment and care in Africa. Human Resources for Health 8, 8-16 (2010). 3. M. Zwahlen, M. Egger, C.-. Bern, Progression and mortality of untreated HIV- positive individuals living in resource-limited settings : Update of literature review and evidence synthesis. UNAIDS Obligation, 1-17. 4. S. C. Keswani, C. A. Pardo, C. L. Cherry, A. Hoke, J. C. Mcarthur, HIV- associated sensory neuropathies. 2105-2117 (2002). 5. D. R. Cornblath, J. C. McArthur, Predominantly sensory neuropathy in patients with AIDS and AIDS‐related complex. Neurology 38, 794-794 (1988). 6. J. C. M. B. MCARTHUR, B.S., Neurologic Manifestations of AIDS. Medicine 66, 407-437 (1987). 7. J. C. McArthur, B. J. Brew, A. Nath, Neurological complications of HIV infection. Lancet neurology 4, 543-555 (2005)10.1016/S1474-4422(05)70165-4). 8. M. Tagliati, J. Grinnell, J. Godbold, D. M. Simpson, Peripheral nerve function in HIV infection: clinical, electrophysiologic, and laboratory findings. Archives of neurology 56, 84 (1999). 9. S. G. Dorsey, P. G. Morton, HIV Peripheral Neuropathy. AACN Clinical Issues: Advanced Practice in Acute and Critical Care 17, 30-36 (2006)10.1097/00044067-200601000-00004). 10. J. a. Canter, D. W. Haas, a. R. Kallianpur, M. D. Ritchie, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. G. Murdock, T. Hulgan, The mitochondrial pharmacogenomics of haplogroup T: MTND2*LHON4917G and antiretroviral therapy-associated peripheral neuropathy. The pharmacogenomics journal 8, 71- 77 (2008)10.1038/sj.tpj.6500470). 11. C. L. Cherry, A. Rosenow, J. S. Affandi, J. C. McArthur, S. L. Wesselingh, P. Price, Cytokine genotype suggests a role for inflammation in nucleoside analog- associated sensory neuropathy (NRTI-SN) and predicts an individual's NRTI-SN risk. AIDS research and human retroviruses 24, 117-123 (2008)10.1089/aid.2007.0168). 12. A. R. Kallianpur, T. Hulgan, J. a. Canter, M. D. Ritchie, J. L. Haines, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. W. Haas, Hemochromatosis (HFE) gene mutations and peripheral neuropathy during antiretroviral therapy. AIDS (London, England) 20, 1503-1513 (2006)10.1097/01.aids.0000237366.56864.3c). 13. C. M. Bailey, K. S. Anderson, in Biochimica et biophysica acta. (2010), vol. 1804, pp. 1213-1222.

118

14. P. An, C. a. Winkler, Host genes associated with HIV/AIDS: advances in gene discovery. Trends in genetics : TIG 26, 119-131 (2010)10.1016/j.tig.2010.01.002). 15. C. C. Elbers, K. R. van Eijk, L. Franke, F. Mulder, Y. T. van der Schouw, C. Wijmenga, N. C. Onland-Moret, Using genome-wide pathway analysis to unravel the etiology of complex diseases. Genetic epidemiology 33, 419-431 (2009)10.1002/gepi.20395). 16. S. Purcell, B. Neale, K. Todd-Brown, L. Thomas, M. A. Ferreira, D. Bender, J. Maller, P. Sklar, P. I. de Bakker, M. J. Daly, P. C. Sham, PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81, 559-575 (2007); published online EpubSep (10.1086/519795). 17. A. L. Price, N. J. Patterson, R. M. Plenge, M. E. Weinblatt, N. A. Shadick, D. Reich, Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 38, 904-909 (2006); published online EpubAug (10.1038/ng1847). 18. T. Tanaka, International HapMap project. Nippon Rinsho 63 Suppl 1, 29-34 (2005). 19. O. Delaneau, J. Marchini, J. F. Zagury, A linear complexity phasing method for thousands of genomes. Nat Methods 9, 179-181 (2012); published online EpubFeb (10.1038/nmeth.1785). 20. B. N. Howie, P. Donnelly, J. Marchini, A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5, e1000529 (2009); published online EpubJun (10.1371/journal.pgen.1000529). 21. L. Southam, K. Panoutsopoulou, N. W. Rayner, K. Chapman, C. Durrant, T. Ferreira, N. Arden, A. Carr, P. Deloukas, M. Doherty, J. Loughlin, A. McCaskie, W. E. Ollier, S. Ralston, T. D. Spector, A. M. Valdes, G. A. Wallis, J. M. Wilkinson, J. Marchini, E. Zeggini, a. consortium, The effect of genome-wide association scan quality control on imputation outcome for common variants. Eur J Hum Genet 19, 610-614 (2011); published online EpubMay (10.1038/ejhg.2010.242). 22. R. C. Team, R. F. f. S. Computing, R: A Language and Environment for Statistical Computing. (Vienna, Austria, 2012). 23. O. A. Panagiotou, J. P. Ioannidis, What should the genome-wide significance threshold be? Empirical replication of borderline genetic associations. International journal of epidemiology 41, 273-286 (2012). 24. G. Schwarzer. (2013). 25. A. Z. Dayem Ullah, N. R. Lemoine, C. Chelala, A practical guide for the functional annotation of genetic variations using SNPnexus. Brief Bioinform 14, 437-447 (2013); published online EpubJul (10.1093/bib/bbt004).

119

26. D. Karolchik, G. P. Barber, J. Casper, H. Clawson, M. S. Cline, M. Diekhans, T. R. Dreszer, P. A. Fujita, L. Guruvadoo, M. Haeussler, R. A. Harte, S. Heitner, A. S. Hinrichs, K. Learned, B. T. Lee, C. H. Li, B. J. Raney, B. Rhead, K. R. Rosenbloom, C. A. Sloan, M. L. Speir, A. S. Zweig, D. Haussler, R. M. Kuhn, W. J. Kent, The UCSC Genome Browser database: 2014 update. Nucleic Acids Res 42, D764-770 (2014); published online EpubJan (10.1093/nar/gkt1168). 27. T. P. Yang, C. Beazley, S. B. Montgomery, A. S. Dimas, M. Gutierrez-Arcelus, B. E. Stranger, P. Deloukas, E. T. Dermitzakis, Genevar: a database and Java application for the analysis and visualization of SNP-gene associations in eQTL studies. Bioinformatics 26, 2474-2476 (2010); published online EpubOct (10.1093/bioinformatics/btq452). 28. L. D. Ward, M. Kellis, HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Research 40, D930-934 (2011)10.1093/nar/gkr917). 29. E. Feingold, P. Good, Guyer, S. Kamholz, L. Liefer, K. Wetterstrand, F. Collins, T. Gingeras, D. Kampa, E. Sekinger, J. Cheng, H. Hirsch, S. Ghosh, Z. Zhu, S. Pate, A. Piccolboni, A. Yang, H. Tammana, S. Bekiranov, P. Kapranov, R. Harrison, G. Church, K. Struhl, B. Ren, T. Kim, L. Barrera, C. Qu, S. Van Calcar, R. Luna, C. Glass, M. Rosenfeld, R. Guigo, S. Antonarakis, E. Birney, M. Brent, L. Pachter, A. Reymond, E. Dermitzakis, C. Dewey, D. Keefe, F. Denoeud, J. Lagarde, J. Ashurst, T. Hubbard, J. Wesselink, R. Castelo, E. Eyras, R. Myers, A. Sidow, S. Batzoglou, N. Trinklein, S. Hartman, S. Aldred, E. Anton, D. Schroeder, S. Marticke, L. Nguyen, J. Schmutz, J. Grimwood, M. Dickson, G. Cooper, E. Stone, G. Asimenos, M. Brudno, A. Dutta, N. Karnani, C. Taylor, H. Kim, G. Robins, G. Stamatoyannopoulos, J. Stamatoyannopoulos, M. Dorschner, P. Sabo, M. Hawrytycz, R. Humbert, J. Wallace, M. Yu, P. Navas, M. McArthur, W. Noble, I. Dunham, C. Koch, R. Andrews, G. Clelland, S. Wilcox, J. Fowler, K. James, P. Groth, O. Dovey, P. Ellis, V. Wraight, A. Mungall, P. Dhami, H. Fiegler, C. Langford, N. Carter, D. Vetrie, M. Snyder, G. Euskirchen, A. Urban, U. Nagalakshmi, J. Rinn, G. Popescu, P. Bertone, S. Hartman, J. Rozowsky, O. Emanuelsson, T. Royce, S. Chung, M. Gerstein, Z. Lian, J. Lian, Y. Nakayama, S. Weissman, V. Stoic, W. Tongprasit, H. Sethi, S. Jones, M. Marra, H. Shin, J. Schein, M. Clamp, K. Lindblad-Toh, J. Chang, D. Jaffe, E. Kamal, E. Lander, T. Mikkelsen, J. Vinson, M. Zody, P. De Jong, K. Osoegawa, M. Nefedov, B. Zhu, A. Baxevanis, T. Wolfsberg, G. Crawford, E. Holt, T. Vasicek, D. Zhou, S. Luo, E. Green, G. Bouffard, E. Margulies, M. Portnoy, N. Hansen, P. Thomas, J. Mcdowell, B. Maskeri, A. Young, Idol, R. Blakesley, G. Schuler, W. Miller, R. Hardison, L. Elnitski, P. Shah, S. Salzberg, M. Pertea, W. Majoros, D. Haussler, D. Thomas, K. Rosenbloom, H. Clawson, A. Siepe, W. Kent, Z. Weng, S. Jin, A. Halees, H. Burden, U. Karaoz, Y. Fu, Y. Yu, C. Ding, C. Cantor, R. Kingston, J. Dennis, R. Green, M. Singer, T. Richmond, J. Norton, P. Farnham, M. Oberley, Inman, McCormick, H. Kim, C. Middle, M. Pirrung, X. Fu, Y. Kwon, Z. Ye, J. Dekker, T. Tabuchi, N. Gheldof, J. Dostie, S. Harvey, E. P. Consortium, The ENCODE (ENCyclopedia Of DNA Elements) Project. Science 306, 636-640 (2004).

120

30. L. Clarke, X. Zheng-Bradley, R. Smith, E. Kulesha, C. Xiao, I. Toneva, B. Vaughan, D. Preuss, R. Leinonen, M. Shumway, S. Sherry, P. Flicek, The 1000 Genomes Project: data management and community access. Nature Methods 9, 459-462 (2012)10.1038/nmeth.1974). 31. L. Han, D. L. Wong, G. Tsai, Z. Jiang, J. T. Coyle, Promoter analysis of human glutamate carboxypeptidase II. Brain Res 1170, 1-12 (2007); published online EpubSep (10.1016/j.brainres.2007.07.017). 32. H. Xie, J. Guo, J. Wang, F. Wang, H. Zhao, C. Liu, L. Wang, X. Lu, L. Wu, Y. Bao, J. Zou, T. Zhang, B. Niu, Glutamate carboxypeptidase II gene polymorphisms and neural tube defects in a high-risk Chinese population. Metabolic brain disease 27, 59-65 (2012)10.1007/s11011-011-9272-8). 33. V. Eklöf, B. Van Guelpen, J. Hultdin, I. Johansson, G. Hallmans, R. Palmqvist, The reduced folate carrier (RFC1) 80G > A and folate hydrolase 1 (FOLH1) 1561C > T polymorphisms and the risk of colorectal cancer: a nested case- referent study. Scand J Clin Lab Invest 68, 393-401 (2008)10.1080/00365510701805431). 34. P. K. Nicholas, L. Mauceri, A. Slate Ciampa, I. B. Corless, N. Raymond, D. J. Barry, A. Viamonte Ros, Distal sensory polyneuropathy in the context of HIV/AIDS. The Journal of the Association of Nurses in AIDS Care : JANAC 18, 32-40 (2007)10.1016/j.jana.2007.05.003). 35. M. Manzoor, J. Runcie, Folate-responsive neuropathy: report of 10 cases. British medical journal 1, 1176-1178 (1976). 36. H. Koike, T. Hama, Y. Kawagashira, R. Hashimoto, M. Tomita, M. Iijima, G. Sobue, The significance of folate deficiency in alcoholic and nutritional neuropathies: analysis of a case. Nutrition 28, 821-824 (2012); published online EpubJul (10.1016/j.nut.2011.11.022). 37. D. J. Lanska, Chapter 30: historical aspects of the major neurological vitamin deficiency disorders: the water-soluble B vitamins. Handb Clin Neurol 95, 445- 476 (2010)10.1016/S0072-9752(08)02130-1). 38. A. Alani, O. Vincent, A. Adewumi, A. Titilope, E. Onogu, A. Ralph, C. Hab, Plasma folate studies in HIV-positive patients at the Lagos university teaching hospital, Nigeria. Indian journal of sexually transmitted diseases 31, 99-103 (2010)10.4103/0253-7184.74995). 39. L. Balfour, J. N. Spaans, D. Fergusson, H. Huff, E. J. Mills, C. J. la Porte, S. Walmsley, N. Singhal, R. Rosenes, N. Tremblay, M. J. Gill, H. Loemba, B. Conway, A. Rachlis, E. Ralph, M. Loutfy, R. Mallick, R. Moorhouse, D. William Cameron, Micronutrient Deficiency and Treatment Adherence in a Randomized Controlled Trial of Micronutrient Supplementation in ART-Naïve Persons with HIV. PLoS One 9, e85607 (2014); published online EpubJan (10.1371/journal.pone.0085607). 40. A. Hazra, K. Wu, P. Kraft, C. S. Fuchs, E. L. Giovannucci, D. J. Hunter, Twenty- four non-synonymous polymorphisms in the one-carbon metabolic pathway and

121

risk of colorectal adenoma in the Nurses' Health Study. Carcinogenesis 28, 1510- 1519 (2007)10.1093/carcin/bgm062). 41. S. Divyya, S. M. Naushad, A. Addlagatta, P. V. Murthy, C. R. Reddy, R. R. Digumarti, S. R. Gottumukkala, S. A. Subbarao, V. K. Kutala, Association of glutamate carboxypeptidase II (GCPII) haplotypes with breast and prostate cancer risk. Gene 516, 76-81 (2013); published online EpubMar (10.1016/j.gene.2012.11.047). 42. K. Chopra, V. Tiwari, Alcoholic neuropathy: possible mechanisms and future treatment possibilities. Br J Clin Pharmacol 73, 348-362 (2012); published online EpubMar (10.1111/j.1365-2125.2011.04111.x). 43. M. Varela-Rey, A. Woodhoo, M. L. Martinez-Chantar, J. M. Mato, S. C. Lu, Alcohol, DNA methylation, and cancer. Alcohol Res 35, 25-35 (2013). 44. P. a. Kandiah, M. Atadzhanov, M. P. Kvalsund, G. L. Birbeck, Evaluating the diagnostic capacity of a single-question neuropathy screen (SQNS) in HIV positive Zambian adults. Journal of neurology, neurosurgery, and psychiatry 81, 1380-1381 (2010)10.1136/jnnp.2009.183210).

122

Chapter 5

5. Genetic Predictors of NRTI Induced Peripheral Neuropathy in Ugandan HIV-1+ Subjects

5.1 Abstract

Sensory peripheral neuropathy is one of the most common toxicities associated with the use of nucleoside reverse transcriptase inhibitors (NRTI-SN), a primary component in antiretroviral therapy. While the role that host genetics plays in the development of

NTRI-SN has been investigated in candidate gene studies, a genome-wide association study to identify genetic predictors of NRTI-SN has not been reported. The objective of our study is to identify genetic predictors of NRTI-SN in a treatment naive Ugandan

HIV+ population. Genotype and phenotype data were collected from 580 Ugandan

HIV+ patients enrolled in a treatment outcome cohort study. Patients received a treatment regimen consisting of zidovudine/lamivudine or stavudine/lamivudine combinations along with either nevirapine or efavirenz. Whole genome genotyping was performed using the Illumina OmniExpress® platform with 681,315 SNPs and imputation with IMPUTE2 resulted in >16 million SNPs available for analysis. Sensory PN was scored on a Likert scale from symptom recording at baseline and each quarterly clinic visit. Associations between genetic markers and NRTI-SN were performed using an additive genetic model and logistic regression. Analyses identified several SNPs associated with NRTI-SN at a genome-wide suggestive significance level (p < 5 x 10-7), including an intergenic SNP, rs188298690 (p = 1.47 x 10-7, OR = 8.61, 95% confidence interval = 20.80 - 3.56), that may influence VAMP4 expression, a protein that regulates asynchronous neurotransmitter release. A candidate gene subanalysis also identified

123

SNPs in two genes associated with NRTI-SN: SNPs in ABCC4 (rs7317112: unadjusted p = 2.8 x 10-3, OR = 0.54, 95% confidence interval = 0.80 - 0.36) and SLC28A1

(rs2242046: unadjusted p = 3.1 x 10-3, OR = 0.19, 95% confidence interval =0.57 –

0.06). These genes are drug transporters and are important for NRTI disposition.

These studies suggest that genetic variation in novel genes involved in nerve function and drug transport may influence an individual patient’s risk of developing NRTI-SN.

Further studies are warranted to investigate the roles these genes play in the development of NRTI-SN.

124

5.2 Introduction

Great progress has been made in the treatment of HIV since the beginning of the

HIV epidemic in the 1980s. HIV infections are treated with Highly Active Antiretroviral

Therapy (HAART) which consists of a combination of drugs that target different stages in the HIV life cycle1,2. One of the most commonly used drug combinations consists of a backbone of two nucleoside reverse transcriptase inhibitors (NRTIs) and one non- nucleoside reverse transcriptase inhibitor (NNRTI) or protease inhibitor (PI)1,2. While

HAART is extremely effective at controlling HIV replication, there are drug toxicities associated with all the classes of drugs used to treat HIV1-3. NRTIs have several toxicities associated with their use including anemia, sensory peripheral neuropathy and renal toxicity4-9. Specifically, the first and second generation NRTIs, including azidothymidine (AZT) and stavudine (d4T) are associated with the development of NRTI induced sensory neuropathies (NRTI-SN). While AZT and d4T are not first line treatments in the developing world, they are still critical to the treatment of HIV in the developing world, particularly in sub-Saharan Africa.

Since the earliest days of HAART, cases of NRTI-SN have been observed, especially when used in combination therapies10,11. NRTI-SN is characterized by the development of numbness, tingling or pain in a distinctive “glove and sock” pattern12.

NRTI-SN is clinically indistinguishable from neuropathies caused by other factors such as HIV infection, diabetes or alcoholism10,11. Unlike other drug toxicities that are dose dependent, the risk of developing NRTI toxicity is highest during the first year of HAART and is reduced after this point13,14. NRTI-SN has been attributed to increases in mitochondrial toxicity mediated by the inhibition of the mitochondrial polymerase,

125 polymerase γ, which is encoded by POLG6,15,16. The inhibition of polymerase γ causes increases in oxidative stress that eventually leads to cellular apoptosis and Wallerian degeneration of peripheral nerves10,17,18. Like other neurodegenerative conditions,

NRTI-SN has also been associated with dysregulation of iron metabolism19.

Multiple genetic studies focusing on genes involved in mitochondrial function or iron metabolism have been conducted, some with positive results. Specifically, genetic mutations in the mitochondrial haplogroup T and the genes HFE and POLG have both been associated with the development of NRTI-SN19-22. While these studies have helped to characterize the role that host genetics play in the development of NRTI-SN, no study to date has examined the whole genome for potential genetic predictors of the development of NRTI-SN.

5.3 Materials and Methods:

5.3.1 Participants

The participants of this study are fully described in Chapter 4.3.1. Genotype and phenotype data from Ugandan HIV+ individuals were collected for the discovery study

(n = 638) and the replication study (n = 209).

5.3.2 Genotyping

The full details of the sample collection and processing, genotyping and quality control procedures for samples and single nucleotide polymorphisms (SNPs) are provided in Chapter 4.3.2. In the discovery study, genomic DNA from 638 subjects was extracted from either saliva or blood samples using standard DNA extraction or

126 normalization techniques. Out of 645 samples (638 patient samples with seven control samples) a total of 585 subjects passed genotyping quality control procedures. After genotyping and imputation quality control procedures 16.9 million SNPs were available for analysis23.

For the replication studies, a total of 169 patients had complete genotype and phenotype data. Genotyping for the replication study was performed according to the manufacturer’s instructions using the Taqman Genotyping assay (Life Technologies,

Grand Island, NY) outlined in Table 5.1.

Table 5.1. SNPs Selected for Replication SNP Assay ID Tag SNPs (r2 >0.8) Custom: rs144134647, rs139815631, rs144690537, rs188298690 AHHS28B rs138815589, rs141776039 Taqman assay IDs are listed. SNPs that are in LD (r2 >0.8) in the LWK 1000 Genomes database are also listed.

5.3.3 Phenotype

Peripheral neuropathy data was gathered quarterly during the study visits as a component of the symptom questionnaire. Peripheral neuropathy is graded on a Likert- type scale with subjects asked if they are experiencing “pain, numbness or tingling in the hands or feet.” A score of “0” denotes no symptoms, “1” means “bothers me not at all”, “2” corresponds to “bothers me a little”, “3” corresponds to “bothers me a moderate amount” and “4” denotes “bothers me a lot”. Case and control status were evaluated from baseline to 12 months from the date of treatment initiation. Subjects that reported

“0” (n = 148) or “1” (n = 31) during the study period or had a decrease in symptoms from baseline (n = 91) were classified as “controls” (n = 270). Subjects that reported “0” or

“1” at baseline and had at least 1 value greater than or equal to “2” during the study

127 period (n = 83), reported “2” at baseline and had at least one value greater than or equal to “3” during the study period (n = 11) or reported “3” at baseline and had at least one value greater than or equal to “4” during the study period (n = 9) were classified as

“cases” (n = 103). A graphical representation of this can be seen in Figure 5.1.

Figure 5.1. NRTI-SN case/control definition method.

5.3.4 Statistical Analyses

Demographic variables were tested for statistical significance using linear regression for continuous variables and ANOVA tests for categorical variables in R24. Standard case/control analyses using logistic regression were performed for the primary and replication analyses using PLINK to test the association between each SNP and the phenotype25. Odds ratios (OR), 95% confidence intervals (95% CI) and p-values were reported for each SNP. To account for the multiple testing burden genome-wide significance was considered p ≤ 5x10-8 and genome-wide suggestive was considered p

≤ 5x10-7; ‘promising’ SNPs at p ≤ 5x10-6 were considered for further bioinformatic

128 analysis26. The control group minor allele frequency was examined for Hardy-Weinberg

Equilibrium for SNPs of interest (Χ2 test, α = 0.05). Linkage disequilibrium calculations were performed in PLINK25. Plots were produced in R and Microsoft Excel 201024.

Meta analyses to combine p-values from the discovery and replication studies were performed in R using the meta package27.

5.3.5 Candidate Gene Analysis

Candidate genes (n = 16) were selected based on literature documenting their function in the metabolic, pharmacokinetic and pharmacodynamic pathways of

NRTIs7,14,19,28-30. Table 5.2 describes the candidate genes selected for this study and their role in NRTI disposition, pharmacology or toxicity. Statistical analyses were carried out in the same fashion as the genome-wide analysis with the exception that the p-value cutoff for significance was determined by performing a Bonferroni correction for the number of haplotype blocks tested (n = 121). SNPs were considered significant if the adjusted p < 0.05. The number of linkage disequilibrium blocks present in each gene was calculated in Haploview using the LWK population31. A gene based cutoff was also considered at p ≤ 0.003 using the number of genes (n = 16) for the multiple testing comparison.

129

Table 5.2. Candidate Genes with NRTI Transport or Functional Evidence

Gene Protein Function/Evidence ABCC4 MRP4 Nucleoside transport32 ABCG2 MXR Drug transport33 HFE HFE Iron metabolism19 POLG POLG Mitochondrial polymerase22 SLC22A6 OAT1 Organic anion transport34-36 SLC22A7 OAT2 Organic anion transport34-36 SLC22A8 OAT3 Organic anion transport34-36 SLC25A19 DNC Nucleoside transport, mitochondrial SLC28A1 CNT1 Nucleoside transport34-36 SLC28A2 CNT2 Nucleoside transport34-36 SLC28A3 CNT3 Nucleoside transport34-36 SLC29A1 ENT1 Nucleoside transport37 SLC29A2 ENT2 Nucleoside transport37 SLC29A3 ENT3 Nucleoside transport37,38 TK1 TK1 Intracellular nucleoside phosphorylation38 TK2 TK2 Intracellular nucleoside phosphorylation38

5.3.6 Bioinformatic Analyses

To further explore the putative biological significance of SNPs that had ‘promising’ p- values (genome-wide p ≤ 5x10-6, candidate gene p ≤ 0.003) bioinformatic analyses were performed. SNPs were annotated to genes using SNPnexus, SCANdb

(http://www.scandb.org/) and the UCSC genome browser39,40. GTEx

(http://www.broadinstitute.org/gtex/) and GeneVar

(http://www.sanger.ac.uk/resources/software/genevar/) databases were employed to examine the effect of a SNP on gene expression41. To determine any regulatory functions of SNPs of interest, Haploreg and ENCODE databases were employed42,43.

130

5.4 Results

5.4.1 Demographic Data

The demographic characteristics of the discovery and replication cohorts are described in Table 5.3. The two cohorts were similar in the majority of demographic variables assessed with the exception of gender percentages, percent of ARKS participants and case rate. The discrepancy of gender is most likely due to the higher percentage of

ARKS participants, as this cohort has a larger male participation rate. The higher rate of cases may also be due to the number of ARKS participants since this cohort takes place in a hospital setting and patients may have better access to care. None of the demographic variables tested had an effect on case/control status with the exception of gender in the initial cohort (p = 0.013). This effect was not seen in the replication cohort

(p = 0.99). The case percentages in the initial and replication cohorts were 23% and

44%, respectively.

Table 5.3. Patient Demographic Data in the Discovery and Replication Cohorts Discovery Replication

Sample size, n 373 169 Age, years, median ± SD 34 ± 8.3 34 ± 9.3 Gender, n (%M) 127 (34%) 73 (43%) Baseline CD4+ T-cell counts, cells/mm3, median ± SD 137 ± 118 200.0 ± 213 Baseline Viral Load, log(copies/mL), median ± SD 5.13 ± 0.69 5.07 ± 0.67 Case, n (%) 103 (28%) 74 (44%) Cohort, n (%ARKS) 117 (23%) 58 (34%)

131

5.4.2 A loci on chromosome 1 is associated with NRTI-SN

After quality control procedures, 16.9 million imputed and genotyped SNPs were tested for association with NRTI-SN using a standard case/control testing methodology.

Figure 5.2 shows the Manhattan plot where the y-axis is the p-value for each SNP and the x-axis is the genomic position. A Q-Q plot showing the observed versus expected p-values and the genomic inflation factor, λ (1.01, Figure 5.3), indicate that there is no significant population stratification.

132

133

Figure 5.2. Manhattan plot showing the distribution along the human autosomes of -log10 (P values) obtained for SNP association with NRTI-SN cases versus control subjects. The red line (p < 5 x 10-6) marks the cutoff for ‘promising’ SNPs considered for further bioinformatics analysis.

133

Figure 5.3. Observed versus expected p-values for SNP association with NRTI-SN cases versus control subjects. P-values are plotted on a -log base 10 scale. The red line indicates the null distribution. The genomic inflation factor λ was 1.01 and was calculated using R statistical computing software.

134

Only 13 SNPs had p-values < 5 x 10-6. SNPs with the greatest association to HIV-

SN were filtered for linkage disequilibrium (r2 >0.8) with the SNP with the lowest p-value for each LD block being retained highlighted in bold in Table 5.4. The SNP with the most significant p-value was on chromosome 1 (rs188298690, p = 1.47 x 10-6, OR =

8.61, 95% CI = 20.80 - 3.56 ). This SNP is located in an intergenic region between

VAMP4 and METTL13 with DMN3 closely downstream of METTL13 (Figure5.4).

Because this was the top SNP in the GWAS and no SNPs reached genome-wide significance, this locus was chosen for further bioinformatic analysis.

Table 5.4. Top Variants Associated with NRTI-Induced Peripheral Neuropathy

Left Right Chr SNP MAF OR U95CI L95CI P Gene Feature Gene Gene 1 rs188298690 0.04 8.61 20.80 3.56 1.47 x 10-6 - Intergenic VAMP4 DNM3 6 rs2234245 0.11 3.39 5.64 2.04 3.00 x 10-6 TREM1 missense NCR2 TREM4 5 rs116426216 0.04 8.01 19.35 3.32 3.05 x 10-6 - Intergenic BASP1 CDH18 11 rs4755601 0.22 2.57 3.80 1.74 3.16 x 10-6 - Intergenic LRRC4C API5 5 rs114363753 0.04 7.96 19.23 3.30 3.17 x 10-6 - Intergenic BASP1 CDH18 5 rs115212730 0.04 7.96 19.23 3.30 3.19 x 10-6 - Intergenic BASP1 CDH18 5 rs141776039 0.04 8.58 21.14 3.48 3.25 x 10-6 - Intergenic BASP1 CDH18 5 rs138815589 0.04 8.58 21.14 3.48 3.25 x 10-6 - Intergenic BASP1 CDH18 5 rs144134647 0.04 8.56 21.09 3.47 3.31 x 10-6 - Intergenic BASP1 CDH18 5 rs139815631 0.04 8.56 21.09 3.47 3.31 x 10-6 - Intergenic BASP1 CDH18 5 rs144690537 0.04 8.56 21.09 3.47 3.31 x 10-6 - Intergenic BASP1 CDH18 11 rs7949101 0.23 2.47 3.66 1.67 3.69 x 10-6 - Intergenic LRRC4C API5 5 rs75759888 0.04 7.37 17.46 3.11 4.67 x 10-6 - Intergenic BASP1 CDH18 The lowest p-value SNP for each LD block is highlighted in bold. Upper 95% Confidence interval, U95CI; Lower 95% Confidence interval, L95CI.

135

136

Figure 5.4. UCSC genome browser image showing position of SNPs associated with NRTI-SN susceptibility in the vicinity of the VAMP4 gene. SNPs in the NRTI CHR1 region track were marginally associated with HIV-SN case vs. control status (p < 1 x 10-5). SNPs in the GeneVar eQTL track are SNPs significantly associated (p < 1 x 10-3) with VAMP4 expression in the LWK Hapmap population. The red box highlights the eQTL region that is unique to the LWK population.

136

5.4.3 The rs188298690 polymorphism is in an active regulatory region and is

located within a VAMP4 eQTL

All of the genes proximal to rs188298690 were investigated to determine if this SNP could regulate their expression. According to Haploreg, rs188298690 is in a region that has active regulatory elements; specifically, it is a weak enhancer in a leukemia cell line,

K562. SNPs that are in LD (r2 >0.8) also change several transcription factor binding motifs. Figure 5.5 shows the genomic position of SNPs in LD with rs188298690 and where they overlap transcription factor binding sites in K562 cells.

Figure 5.5. SNPs that are in LD with rs188298690 are in regions that have active regulatory elements. The “User Supplied Track” plots SNPs that are in LD (r2 > 0.8) with rs188298690. The red boxes highlight the alignment of SNPs with H3K27Ac marks, DNase1 hypersensitivity clusters and transcription factor binding sites in the ENCODE data.

Unfortunately, rs188298690 and the SNPs in LD with it are not present in the GTEx database, so tissue specific expression analyses could not be performed. However, eQTL data in the GeneVar database was available for VAMP4, which allowed an

137 examination of the eQTL expression pattern of VAMP4 around the rs188298690 locus.

The eQTL pattern in LCL cells was examined for three African Hapmap populations:

Luhya in Webuye, Kenya (LWK), Maasai in Kinyawa, Kenya (MKK) and Yoruba in

Ibadan, Nigeria (YRI). The study population is most genetically similar to the LWK population. This examination revealed that the LWK population has a unique additional eQTL locus for VAMP4 which is located farther upstream from the eQTL loci that is not seen in the other populations (Figure 5.6).

Figure 5.6. Genevar VAMP4 eQTL LCL data for three African populations. The LWK population possesses a unique eQTL locus (highlighted in the green box) that is associated with VAMP4 expression. LWK = Luhya in Webuye, Kenya, MKK = Maasai in Kinyawa, Kenya, YRI = Yoruba in Ibadan, Nigeria

138

rs188298690 had a minor allele frequency of 10% and 1% in case and control groups, respectively (Figure 5.7). The control group minor allele frequency of 1% was comparable to the 1000 Genomes African population minor allele frequency for this

SNP of 1%44. Expression of METTL13 and DNM2 was not associated with the rs188298690 locus.

0.25

0.2

0.15

0.1

0.1 Minor Allele MinorAllele Frequency

0.05

0.01

0 Controls Cases Case/Control Status

Figure 5.7. The association of intergenic SNP rs188298690 with NRTI-SN case and control status in HIV+ Ugandan subjects.

139

Minor allele frequencies of rs188298690 in the discovery cohort are significantly higher (p = 1.47 x 10-6) in case subjects (n=103) versus control subjects (n=270).

5.4.4 Replication Results

A cohort of 169 new subjects from the original UARTO and ARKS cohorts were available for the replication study (Table 5.5). The top SNP from the GWAS study was chosen for replication and genotyped using Taqman genotyping assays. A meta- analysis was also performed to see the overall effect of this SNP in both the discovery and replication cohorts. While the SNP did not replicate it did maintain odds ratio directionality.

140

Table 5.5. Results from Replication and Meta-Analysis

Discovery Replication Meta-analysis

Minor U96 L96 U96 L96 U96 L96 SNP Chr OR P OR P OR P* Allele CI CI CI CI CI CI

-6 -5 rs188298690 1 A 8.61 20.8 3.56 1.47 x 10 1.59 3.78 0.67 0.29 3.67 6.80 1.98 1.8 x 10

* Meta-analysis p-values reflect the one-sided p-value because the effect direction was assumed a priori to be the same direction in the replication cohort as the discovery cohort. Upper 95% Confidence interval, U95CI; Lower 95% Confidence interval, L95CI.

141

141

5.4.5 Candidate gene study reveals an association of ABCC4 polymorphisms

and NRTI-SN

In a candidate gene subanalysis of 16 genes, only three SNPs had p-values reaching the gene-based significance cutoff of p < 3 x 10-3 (Table 5.6) and no SNPs met the haplotype-based significance cutoff of adjusted p < 0.05. The top two unadjusted p- value SNPs are in strong LD (r2 = 1) in the LWK Hapmap population and are located in the first intron of the ABCC4 gene, which encodes the MRP4 drug transporter

(rs7317112: unadjusted p = 2.8 x 10-3, OR = 0.54, 95% CI = 0.80 - 0.36; rs8001475: unadjusted p = 2.9 x 10-3, OR = 0.55, 95% CI = 0.81 - 0.37) (Figure 5.8). The third most associated SNP is a missense mutation in SLC28A1, which encodes the CNT1 drug transporter (rs2242046: unadjusted p = 3.1 x 10-3, OR = 0.19, 95% CI = 0.57 - 0.06).

These SNPs were chosen for further bioinformatic analysis.

Table 5.6. Top Candidate Gene Variants Associated with NRTI-Induced Peripheral Neuropathy U95 L95 P SNP MAF OR P Gene Feature CI CI adjusted* rs7317112 0.46 0.54 0.80 0.36 0.0028 0.34 ABCC4 Intronic rs8001475 0.46 0.55 0.81 0.37 0.0029 0.35 ABCC4 Intronic rs2242046 0.08 0.19 0.57 0.06 0.0031 0.38 SLC28A1 Missense *P values adjusted for the number of haplotype blocks tested across genes. The lowest p-value SNP for each LD block is highlighted in bold. Upper 95% Confidence interval, U95CI; Lower 95% Confidence interval, L95CI.

142

143

Figure 5.8. UCSC genome browser image showing position of the top candidate SNP (rs7317112, in purple) associated with NRTI- -3 SN susceptibility in the vicinity of ABCC4 (p = 2.8 x 10 ).

143

Haploreg indicates that rs7317112 is highly evolutionarily conserved and located within the first intron of ABCC4 with multiple active regulatory elements that function as a weak enhancer in multiple tissues. In a GTEx analysis, rs7317112 is associated with increased RNA expression (p = 0.02; Figure 5.9). According to Haploreg, rs2242046 is an enhancer; however GTEx bioinformatic analyses did not reveal any association with the SNP and SLC28A1 RNA expression.

Figure 5.9. The effect of ABCC4 SNP rs7317112 genotype on RNA expression in nerve tissue. ABCC4 levels increase significantly with the number of variant alleles present in the genotype (n = 97, p < 0.02). Expression is from RNA-seq (log[RPKM]) and was provided by the GTEx bioinformatic database. RPKM = reads per kilobase per million.

144

Figure 10 5.5 Discussion

NRTI-SN is a commonly observed side effect of HAART therapy10. While some risk factors and genetic predictors of NRTI-SN are known, there still is uncertainty about what genes have the most influence on the development of NRTI-SN. This study used a whole genome and candidate gene approach to discover novel genetic predictors of

NRTI-SN. While no SNPs reached genome-wide significance, the top hit

(rs188829890) had a ‘promising’ p-value (p = 1.47 x 10-6). This SNP was found to be in a LWK population specific eQTL locus for VAMP4. The candidate gene study revealed three SNPs in the ABCC4 (rs731112, unadjusted p = 2.8 x 10-3; rs8001475, unadjusted p = 2.9 x 10-3) and SLC28A1 (rs2242046, unadjusted p = 3.1 x 10-3) genes that reached gene-based significance (p < 3 x 10-3). No SNPs reached haplotype-corrected significance. Publically available data shows that rs731112 is associated with increased

ABCC4 expression in nerve tissue and is in an active regulatory region. Neither of these

SNPs reached statistical significance in the replication study; however this may be due to the small sample size of the replication cohort. rs731112 had a slightly improved p- value in the meta-analysis. Despite the limited power of these studies, the results for these genome-wide and candidate gene studies are worthy of follow-up studies.

VAMP4 is a largely uncharacterized gene that is in the vesicle-associated membrane protein (VAMP)/synaptobrevin family. These proteins generally are involved in the docking and/or fusion of synaptic vesicles with the presynaptic membrane.

VAMP4, unlike other VAMPs, has been shown to selectively maintain bulk Ca2+- dependent asynchronous release in neuronal cells45. When the function of VAMP4 is

145 reduced, nerves continue conducting without a pause between neurotransmitter releases45. The role that VAMP4 may play in the development of NRTI-SN is still unclear, however it can be surmised that a reduction of VAMP4 would lead to increases in presynaptic neurotransmitter release, which will lead to increases in stimulation of the post-synaptic neuron. An increase of stimulation of the postsynaptic neuron leads to increases in the expression of several molecular signaling molecules associated with neuropathic pain including cytokines, COX2 enzymes and ion channels46. NRTIs are known to cause neuronal toxicity by causing mitochondrial damage7,28,47. This is the same mechanism that is seen in some inherited neuropathies such as Charcot-Marie

Tooth disease48. Additionally, changes in neurotransmitter release have been documented to be one mechanism of neuropathic pain49. Therefore it is possible that after a neuron has been damaged by NRTI exposure, a decrease in VAMP4 expression could lead to increases in neuropathic pain due to a lack of regulation of neurotransmitter release.

CNT1 (SLC28A1) and MRP4 (ABCC4) can transport NRTIs50,51,52. CNT1 is an influx transporter and MRP4 is an efflux transporter; both of these transporters should influence NRTI systemic exposure and possibly the levels of NRTI in the dorsal root ganglion. rs7317112 is associated with increased expression of ABCC4 in nerve tissue.

As MRP4 is an efflux transporter, increases in expression of this protein would lead to a reduced amount of drug accumulating in the nerve cell, which would decrease the neuronal damage caused by NRTI exposure, consistent with the protective effect observed in the current study. This SNP has also been shown to predict methotrexate plasma levels in pediatric acute lymphoblastic leukemia patients53. SLC28A1 is

146 primarily apically expressed in the liver and kidney and importantly is expressed in the dorsal root ganglion 54,55. rs2242046 causes a missense mutation in the SLC28A1 gene and would be expected to result in less systemic drug exposure and less drug entering the nerve cell and therefore less exposure to NRTIs. While this SNP has not been functionally characterized, it has been shown to predict clinical outcomes in breast cancer patients receiving gemcitabine which is also a nucleoside analog56. While other studies have seen an association with HFE and POLG and NRTI-SN, these findings have not been replicated and this study was also not able to replicate these findings19,22. Other studies have also identified mutations in mitochondrial DNA

(mtDNA) that are associated with NRTI-SN, however, mtDNA was not available in this study20,21,57.

The largest limitation to this study is a lack of power to detect small effect sizes due to a small sample size. This is evident by the lack of SNPs that met genome-wide significance. However, this study is meant to be hypothesis generating and provides novel genes for further investigation. Another potentially confounding factor is neuropathy caused by HIV infection. Although an effort was made to account for peripheral neuropathy due to HIV infection, there is a possibility that case status was assigned incorrectly. Additionally, the definition of the phenotype may be skewed by the patients’ perception of pain due to advanced HIV infection. To attenuate this effect a clinician that specializes in HIV patients was consulted to ensure proper phenotype definition.

147

5.6 Conclusion

This study identified several potential genetic predictors for the development of

NRTI-SN in genes with known roles in nerve function or intracellular NRTI exposure.

Further in vitro and clinical studies are warranted to define a role for VAMP4, MRP4,

CNT and OAT1 in NRTI-induced peripheral neuropathy. A long term goal is to define genetic markers that could be used to identify patients with modified risk for NRTI-SN prior to drug exposure. Genetic studies such as these may also identify novel targets for treating peripheral neuropathies including, diabetic neuropathy, chemotherapeutic neuropathy and congenital neuropathies.

148

5.7 References

1. W. H. Organization. (2013). 2. P. o. A. G. f. A. a. Adolescents, Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Department of Health and Human Services, (2013). 3. E. De Clercq, The nucleoside reverse transcriptase inhibitors, nonnucleoside reverse transcriptase inhibitors, and protease inhibitors in the treatment of HIV infections (AIDS). Advances in pharmacology (San Diego, Calif.) 67, 317-358 (2013)10.1016/B978-0-12-405880-4.00009-3). 4. C. Foster, H. Lyall, HIV and mitochondrial toxicity in children. The Journal of antimicrobial chemotherapy 61, 8-12 (2008)10.1093/jac/dkm411). 5. T. N. Kakuda, Pharmacology of nucleoside and nucleotide reverse transcriptase inhibitor-induced mitochondrial toxicity. Clinical therapeutics 22, 685-708 (2000)10.1016/S0149-2918(00)90004-3). 6. J. J. Kohler, W. Lewis, A Brief Overview of Mechanisms of Mitochondrial Toxicity From NRTIs. Journal of Clinical Virology 172, 166-172 (2007)10.1002/em). 7. B. Liang, Æ. D. J. Kleinhenz, Æ. E. R. Walp, Æ. S. Dikalov, D. P. Jones, Æ. R. F. Schinazi, Æ. R. L. Sutliff, E. R. Kline, L. Bassit, B. I. Hernandez-Santiago, M. a. Detorio, D. J. Kleinhenz, E. R. Walp, S. Dikalov, R. F. Schinazi, R. L. Sutliff, Long-term exposure to AZT, but not d4T, increases endothelial cell oxidative stress and mitochondrial dysfunction. Cardiovascular toxicology 9, 1-12 (2009)10.1007/s12012-008-9029-8). 8. N. Venhoff, B. Setzer, K. Melkaoui, U. a. Walker, Mitochondrial toxicity of tenofovir, emtricitabine and abacavir alone and in combination with additional nucleoside reverse transcriptase inhibitors. Antiviral therapy 12, 1075-1085 (2007). 9. J. J. Kohler, S. H. Hosseini, A. Hoying-Brandt, E. Green, D. M. Johnson, R. Russ, D. Tran, C. M. Raper, R. Santoianni, W. Lewis, Tenofovir renal toxicity targets mitochondria of renal proximal tubules. Laboratory investigation; a journal of technical methods and pathology 89, 513-519 (2009)10.1038/labinvest.2009.14). 10. S. C. Keswani, C. A. Pardo, C. L. Cherry, A. Hoke, J. C. Mcarthur, HIV- associated sensory neuropathies. 2105-2117 (2002). 11. P. K. Nicholas, L. Mauceri, A. Slate Ciampa, I. B. Corless, N. Raymond, D. J. Barry, A. Viamonte Ros, Distal sensory polyneuropathy in the context of HIV/AIDS. The Journal of the Association of Nurses in AIDS Care : JANAC 18, 32-40 (2007)10.1016/j.jana.2007.05.003). 12. D. T. Dieterich, Long-term complications of nucleoside reverse transcriptase inhibitor therapy. AIDS Read 13, 176-184, 187 (2003); published online EpubApr (

149

13. A. ARENAS-PINTO, K. Bhaskaran, D. Dunn, I. V. D. Weller, L. V. D. WELLER, The risk of developing peripheral neuropathy induced by nucleoside reverse transcriptase inhibitors decreases over time : evidence from the Delta trial. Antiviral therapy 13, 289-295. 14. P. L. Anderson, T. N. Kakuda, K. A. Lichtenstein, The cellular pharmacology of nucleoside- and nucleotide-analogue reverse-transcriptase inhibitors and its relationship to clinical toxicities. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 38, 743-753 (2004)10.1086/381678). 15. W. Lewis, B. J. Day, W. C. Copeland, Mitochondrial toxicity of NRTI antiviral drugs: an integrated cellular perspective. Nature reviews. Drug discovery 2, 812- 822 (2003)10.1038/nrd1201). 16. W. Lewis, Pharmacogenomics, toxicogenomics, and DNA polymerase gamma. The Journal of infectious diseases 195, 1399-1401 (2007)10.1086/513879). 17. J. C. McArthur, B. J. Brew, A. Nath, Neurological complications of HIV infection. Lancet neurology 4, 543-555 (2005)10.1016/S1474-4422(05)70165-4). 18. A. R. Kallianpur, Pharmacogenetics of nucleoside peripheral neuropathy R eview. In Vivo 10, 623-637 (2009). 19. A. R. Kallianpur, T. Hulgan, J. a. Canter, M. D. Ritchie, J. L. Haines, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. W. Haas, Hemochromatosis (HFE) gene mutations and peripheral neuropathy during antiretroviral therapy. AIDS (London, England) 20, 1503-1513 (2006)10.1097/01.aids.0000237366.56864.3c). 20. J. a. Canter, D. W. Haas, a. R. Kallianpur, M. D. Ritchie, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. G. Murdock, T. Hulgan, The mitochondrial pharmacogenomics of haplogroup T: MTND2*LHON4917G and antiretroviral therapy-associated peripheral neuropathy. The pharmacogenomics journal 8, 71- 77 (2008)10.1038/sj.tpj.6500470). 21. T. Hulgan, D. W. Haas, J. L. Haines, M. D. Ritchie, G. K. Robbins, R. W. Shafer, D. B. Clifford, A. R. Kallianpur, M. Summar, J. a. Canter, Mitochondrial haplogroups and peripheral neuropathy during antiretroviral therapy: an adult AIDS clinical trials group study. AIDS (London, England) 19, 1341-1349 (2005). 22. H. Yamanaka, H. Gatanaga, P. Kosalaraksa, S. Matsuoka-Aizawa, T. Takahashi, S. Kimura, S. Oka, Novel mutation of human DNA polymerase gamma associated with mitochondrial toxicity induced by anti-HIV treatment. The Journal of infectious diseases 195, 1419-1425 (2007)10.1086/513872). 23. B. N. Howie, P. Donnelly, J. Marchini, A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5, e1000529 (2009); published online EpubJun (10.1371/journal.pgen.1000529). 24. R. C. Team, R. F. f. S. Computing, R: A Language and Environment for Statistical Computing. (Vienna, Austria, 2012).

150

25. S. Purcell, B. Neale, K. Todd-Brown, L. Thomas, M. A. Ferreira, D. Bender, J. Maller, P. Sklar, P. I. de Bakker, M. J. Daly, P. C. Sham, PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81, 559-575 (2007); published online EpubSep (10.1086/519795). 26. O. A. Panagiotou, J. P. Ioannidis, What should the genome-wide significance threshold be? Empirical replication of borderline genetic associations. International journal of epidemiology 41, 273-286 (2012). 27. G. Schwarzer. (2013). 28. C. M. Bailey, K. S. Anderson, in Biochimica et biophysica acta. (2010), vol. 1804, pp. 1213-1222. 29. M. G. Belinsky, P. Guo, K. Lee, F. Zhou, E. Kotova, A. Grinberg, H. Westphal, I. Shchaveleva, A. Klein-Szanto, J. M. Gallo, G. D. Kruh, Multidrug resistance protein 4 protects bone marrow, thymus, spleen, and intestine from nucleotide analogue-induced damage. Cancer research 67, 262-268 (2007)10.1158/0008- 5472.CAN-06-2680). 30. R. J. Bienstock, T. Ludaway, J. Mcnaught, R. Russ, Antiretroviral nucleosides, deoxynucleotide carrier and mitochondrial DNA: evidence supporting the DNA pol γ hypothesis. Environmental Health 20, 675-684 (2007). 31. J. C. Barrett, B. Fry, J. Maller, M. J. Daly, Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21, 263-265 (2005); published online EpubJan (10.1093/bioinformatics/bth457). 32. N. Abla, L. W. Chinn, T. Nakamura, L. Liu, C. C. Huang, S. J. Johns, M. Kawamoto, D. Stryke, T. R. Taylor, T. E. Ferrin, K. M. Giacomini, D. L. Kroetz, The human multidrug resistance protein 4 (MRP4, ABCC4): functional analysis of a highly polymorphic gene. J Pharmacol Exp Ther 325, 859-868 (2008); published online EpubJun (10.1124/jpet.108.136523). 33. X. Wang, M. Baba, The role of breast cancer resistance protein (BCRP/ABCG2) in cellular resistance to HIV-1 nucleoside reverse transcriptase inhibitors. Antivir Chem Chemother 16, 213-216 (2005). 34. E. Errasti-Murugarren, M. Pastor-Anglada, Drug transporter pharmacogenetics in nucleoside based therapies. Pharmacogenomics 11, 809-841 (2010). 35. A. L. VanWert, M. R. Gionfriddo, D. H. Sweet, Organic anion transporters: discovery, pharmacology, regulation and roles in pathophysiology. Biopharm Drug Dispos 31, 1-71 (2010); published online EpubJan (10.1002/bdd.693). 36. T. P. Zimmerman, W. B. Mahony, K. L. Prus, 3'-azido-3'-deoxythymidine. An unusual nucleoside analogue that permeates the membrane of human erythrocytes and lymphocytes by nonfacilitated diffusion. J Biol Chem 262, 5748- 5754 (1987); published online EpubApr ( 37. S. a. Baldwin, P. R. Beal, S. Y. M. Yao, A. E. King, C. E. Cass, J. D. Young, The equilibrative nucleoside transporter family, SLC29. Pflügers Archiv : European journal of physiology 447, 735-743 (2004)10.1007/s00424-003-1103-2).

151

38. G. J. Veal, D. J. Back, Metabolism of Zidovudine. Gen Pharmacol 26, 1469-1475 (1995); published online EpubNov ( 39. A. Z. Dayem Ullah, N. R. Lemoine, C. Chelala, A practical guide for the functional annotation of genetic variations using SNPnexus. Brief Bioinform 14, 437-447 (2013); published online EpubJul (10.1093/bib/bbt004). 40. D. Karolchik, G. P. Barber, J. Casper, H. Clawson, M. S. Cline, M. Diekhans, T. R. Dreszer, P. A. Fujita, L. Guruvadoo, M. Haeussler, R. A. Harte, S. Heitner, A. S. Hinrichs, K. Learned, B. T. Lee, C. H. Li, B. J. Raney, B. Rhead, K. R. Rosenbloom, C. A. Sloan, M. L. Speir, A. S. Zweig, D. Haussler, R. M. Kuhn, W. J. Kent, The UCSC Genome Browser database: 2014 update. Nucleic Acids Res 42, D764-770 (2014); published online EpubJan (10.1093/nar/gkt1168). 41. T. P. Yang, C. Beazley, S. B. Montgomery, A. S. Dimas, M. Gutierrez-Arcelus, B. E. Stranger, P. Deloukas, E. T. Dermitzakis, Genevar: a database and Java application for the analysis and visualization of SNP-gene associations in eQTL studies. Bioinformatics 26, 2474-2476 (2010); published online EpubOct (10.1093/bioinformatics/btq452). 42. L. D. Ward, M. Kellis, HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Research 40, D930-934 (2011)10.1093/nar/gkr917). 43. E. Feingold, P. Good, Guyer, S. Kamholz, L. Liefer, K. Wetterstrand, F. Collins, T. Gingeras, D. Kampa, E. Sekinger, J. Cheng, H. Hirsch, S. Ghosh, Z. Zhu, S. Pate, A. Piccolboni, A. Yang, H. Tammana, S. Bekiranov, P. Kapranov, R. Harrison, G. Church, K. Struhl, B. Ren, T. Kim, L. Barrera, C. Qu, S. Van Calcar, R. Luna, C. Glass, M. Rosenfeld, R. Guigo, S. Antonarakis, E. Birney, M. Brent, L. Pachter, A. Reymond, E. Dermitzakis, C. Dewey, D. Keefe, F. Denoeud, J. Lagarde, J. Ashurst, T. Hubbard, J. Wesselink, R. Castelo, E. Eyras, R. Myers, A. Sidow, S. Batzoglou, N. Trinklein, S. Hartman, S. Aldred, E. Anton, D. Schroeder, S. Marticke, L. Nguyen, J. Schmutz, J. Grimwood, M. Dickson, G. Cooper, E. Stone, G. Asimenos, M. Brudno, A. Dutta, N. Karnani, C. Taylor, H. Kim, G. Robins, G. Stamatoyannopoulos, J. Stamatoyannopoulos, M. Dorschner, P. Sabo, M. Hawrytycz, R. Humbert, J. Wallace, M. Yu, P. Navas, M. McArthur, W. Noble, I. Dunham, C. Koch, R. Andrews, G. Clelland, S. Wilcox, J. Fowler, K. James, P. Groth, O. Dovey, P. Ellis, V. Wraight, A. Mungall, P. Dhami, H. Fiegler, C. Langford, N. Carter, D. Vetrie, M. Snyder, G. Euskirchen, A. Urban, U. Nagalakshmi, J. Rinn, G. Popescu, P. Bertone, S. Hartman, J. Rozowsky, O. Emanuelsson, T. Royce, S. Chung, M. Gerstein, Z. Lian, J. Lian, Y. Nakayama, S. Weissman, V. Stoic, W. Tongprasit, H. Sethi, S. Jones, M. Marra, H. Shin, J. Schein, M. Clamp, K. Lindblad-Toh, J. Chang, D. Jaffe, E. Kamal, E. Lander, T. Mikkelsen, J. Vinson, M. Zody, P. De Jong, K. Osoegawa, M. Nefedov, B. Zhu, A. Baxevanis, T. Wolfsberg, G. Crawford, E. Holt, T. Vasicek, D. Zhou, S. Luo, E. Green, G. Bouffard, E. Margulies, M. Portnoy, N. Hansen, P. Thomas, J. Mcdowell, B. Maskeri, A. Young, Idol, R. Blakesley, G. Schuler, W. Miller, R. Hardison, L. Elnitski, P. Shah, S. Salzberg, M. Pertea, W. Majoros, D. Haussler, D. Thomas, K. Rosenbloom, H. Clawson, A. Siepe, W. Kent, Z. Weng, S. Jin, A.

152

Halees, H. Burden, U. Karaoz, Y. Fu, Y. Yu, C. Ding, C. Cantor, R. Kingston, J. Dennis, R. Green, M. Singer, T. Richmond, J. Norton, P. Farnham, M. Oberley, Inman, McCormick, H. Kim, C. Middle, M. Pirrung, X. Fu, Y. Kwon, Z. Ye, J. Dekker, T. Tabuchi, N. Gheldof, J. Dostie, S. Harvey, E. P. Consortium, The ENCODE (ENCyclopedia Of DNA Elements) Project. Science 306, 636-640 (2004). 44. L. Clarke, X. Zheng-Bradley, R. Smith, E. Kulesha, C. Xiao, I. Toneva, B. Vaughan, D. Preuss, R. Leinonen, M. Shumway, S. Sherry, P. Flicek, The 1000 Genomes Project: data management and community access. Nature Methods 9, 459-462 (2012)10.1038/nmeth.1974). 45. J. Raingo, M. Khvotchev, P. Liu, F. Darios, Y. C. Li, D. M. Ramirez, M. Adachi, P. Lemieux, K. Toth, B. Davletov, E. T. Kavalali, VAMP4 directs synaptic vesicles to a pool that selectively maintains asynchronous neurotransmission. Nat Neurosci 15, 738-745 (2012); published online EpubMay (10.1038/nn.3067). 46. R. Baron, A. Binder, G. Wasner, Neuropathic pain: diagnosis, pathophysiological mechanisms, and treatment. Lancet Neurol 9, 807-819 (2010); published online EpubAug (10.1016/S1474-4422(10)70143-5). 47. D. Höschele, Cell culture models for the investigation of NRTI-induced mitochondrial toxicity. Relevance for the prediction of clinical toxicity. Toxicology in vitro : an international journal published in association with BIBRA 20, 535-546 (2006)10.1016/j.tiv.2005.11.007). 48. M. Sajic, Mitochondrial Dynamics in Peripheral Neuropathies. Antioxid Redox Signal, (2014); published online EpubFeb (10.1089/ars.2013.5822). 49. M. Costigan, J. Scholz, C. J. Woolf, Neuropathic pain: a maladaptive response of the nervous system to damage. Annu Rev Neurosci 32, 1-32 (2009)10.1146/annurev.neuro.051508.135531). 50. C. a. Ritter, G. Jedlitschky, H. Meyer zu Schwabedissen, M. Grube, K. Köck, H. K. Kroemer, Cellular export of drugs and signaling molecules by the ATP-binding cassette transporters MRP4 (ABCC4) and MRP5 (ABCC5). Drug metabolism reviews 37, 253-278 (2005)10.1081/DMR-200047984). 51. L. I. Cano-Soldado P, Molina-Arcas M, Casado FJ, Martinez-Picado J, Lostao MP, Pastor-Anglada M., Interaction of nucleoside inhibitors of HIV-1 reverse transcriptase with the concentrative nucleoside transporter-1 (SLC28A1). Antivir Ther. 9, 993-1002 (2004). 52. D. M. Truong, G. Kaler, A. Khandelwal, P. W. Swaan, S. K. Nigam, Multi-level analysis of organic anion transporters 1, 3, and 6 reveals major differences in structural determinants of antiviral discrimination. The Journal of biological chemistry 283, 8654-8663 (2008)10.1074/jbc.M708615200). 53. E. Lopez-Lopez, J. Ballesteros, M. A. Piñan, J. Sanchez de Toledo, N. Garcia de Andoin, P. Garcia-Miguel, A. Navajas, A. Garcia-Orad, Polymorphisms in the methotrexate transport pathway: a new tool for MTX plasma level prediction in

153

pediatric acute lymphoblastic leukemia. Pharmacogenet Genomics 23, 53-61 (2013); published online EpubFeb (10.1097/FPC.0b013e32835c3b24). 54. M. Pastor-Anglada, P. Cano-Soldado, E. Errasti-Murugarren, F. J. Casado, SLC28 genes and concentrative nucleoside transporter (CNT) proteins. Xenobiotica 38, 972-994 (2008); published online EpubJul (10.1080/00498250802069096). 55. C. Wu, I. Macleod, A. I. Su, BioGPS and MyGene.info: organizing online, gene- centric information. Nucleic Acids Res 41, D561-565 (2013); published online EpubJan (10.1093/nar/gks1114). 56. S. Y. Lee, S. A. Im, Y. H. Park, S. Y. Woo, S. Kim, M. K. Choi, W. Chang, J. S. Ahn, Y. H. Im, Genetic polymorphisms of SLC28A3, SLC29A1 and RRM1 predict clinical outcome in patients with metastatic breast cancer receiving gemcitabine plus paclitaxel chemotherapy. Eur J Cancer 50, 698-705 (2014); published online EpubMar (10.1016/j.ejca.2013.11.028). 57. J. a. Canter, G. K. Robbins, D. Selph, D. B. Clifford, A. R. Kallianpur, R. Shafer, S. Levy, D. G. Murdock, M. D. Ritchie, D. W. Haas, T. Hulgan, African mitochondrial DNA subhaplogroups and peripheral neuropathy during antiretroviral therapy. The Journal of infectious diseases 201, 1703-1707 (2010)10.1086/652419).

154

Chapter 6

6. Conclusions>

Since its emergence in the 1980’s, HIV has been one of the most important epidemics worldwide1. HIV is a lentivirus that affects the immune system by depleting the T-cell population of the host and if left untreated results in death due to opportunistic infections2. In addition to immune system dysfunction, HIV infection also results in numerous complications which may result in wasting, neurologic and other complications3,4. HIV is a genetically diverse virus that includes groups and subgroups that have differing regional prevalences5.

Since the approval of azidothymidine, 28 drugs have been approved for the treatment of HIV and more are currently in the drug development pipeline. HIV treatment generally consists of inhibiting viral replication through multiple mechanisms using highly active antiretroviral therapy (HAART)6-8. In the developing world, HAART consists of drug regimens containing two nucleoside reverse transcriptase inhibitors

(NRTI) and one non-nucleoside reverse transcriptase inhibitor (NNRTI) or a protease inhibitor (PI)7. These drug regimens are efficacious and consistently reduce viral load and increase CD4+ T-cell counts, however, the first and second generation of antiretrovirals (ARV) that are used in the developing world have greater toxicity than their newer counterparts9,10.

There have been numerous pharmacogenomic associations impacting the pharmacokinetics, pharmacodynamics and toxicity of ARVs, many of them in genes important in metabolism and disposition of ARVs11-14. The focus of these studies has

155 been primarily on drug metabolizing enzymes (CYP450s and UGTs) and in membrane transporters in the SLC and ABC superfamilies.

This dissertation describes research to further characterize the role that host genetics play in the development of HIV infection complications, ARV pharmacology and ARV toxicity. Specifically, it is focused on the role of host genetics in HIV-induced peripheral neuropathy (HIV-SN), the pharmacogenetics of nevirapine (NVP) pharmacokinetics and the pharmacogenetics of NRTI-induced peripheral neuropathy

(NRTI-SN). The overall goal of this dissertation is to increase the understanding of the role of host genetics on HIV infection complications and ARV pharmacogenetics and to provide data to direct further research into these fields of study.

In chapters 2 and 3, research was performed to assess and characterize the heritability of NVP pharmacokinetics. It has been shown previously that NVP is metabolized by CYP2B6 and associations between polymorphisms in this gene have been shown to impact NVP pharmacokinetics15-19. The role is unclear for ABCB1, which has been associated with NVP toxicity, but associations between ABCB1 and NVP pharmacokinetics have been controversial13,17,20. While numerous candidate gene studies investigating the effect of genetics on NVP pharmacokinetics have been performed, the heritability of NVP pharmacokinetics is unknown. To address this, in chapter 2, a study to determine the overall relative genetic contribution to the variance in NVP AUC0-6hr was performed and a significant relative genetic contribution was found in European and African American subjects. To investigate the role that polymorphisms in the CYP2B6 and ABCB1 genes play in NVP pharmacokinetics, a study investigating the association between candidate polymorphisms and NVP AUC0-6hr was performed.

156

While no statistically significant associations were found, likely due to the limited sample size, a trend towards association was observed for the CYP2B6 516G>T polymorphism.

To further investigate the results in chapter 2, an additional candidate gene study investigating the genetics of NVP pharmacokinetics was performed in HIV+ Ugandans.

Significant associations between NVP Cmin and previously known polymorphisms in

CYP2B6 and a novel polymorphism in ABCC10 were observed. These results reiterate the importance of CYP2B6 in NVP pharmacokinetics and also provide novel evidence that ABCC10 is also important in NVP pharmacokinetics. Additionally, a composite genotype consisting of these polymorphisms was was a predictor of NVP Cmin. The effect of composite genotypes has been reported for CYP2B6 and efavirenz pharmacokinetics, but not for NVP21. The combined effect of CYP2B6 and ABCC10 suggests that variation in NVP pharmacokinetics is polygenic.

One of the main complications of HIV infection is peripheral neuropathy22. While the development of HIV-SN is rare in the developed world, it occurs with a higher frequency in the developing world because of difficulties in access to healthcare23. The mechanism of HIV-SN is poorly understood and no studies investigating the role that host genetics play in the development of HIV-SN have been performed. In chapter 4, a genome-wide association study with a case vs. control design was used to investigate the role that host genetics play in the development of HIV-SN. A SNP proximal to the 5’ end of

FOLH1, a gene important in folate metabolism, was associated with the development of

HIV-SN, with a higher incidence of the variant allele in HIV-SN cases. This SNP is in a region that has active regulatory features and is associated with a decrease in FOLH1 expression in neuronal tissue. A trend towards an association between the FOLH1

157

SNP and HIV-SN was found in a replication cohort and statistical significance was improved when a meta analysis of the discovery and replication studies was performed.

Folate responsive peripheral neuropathies have been extensively documented, particularly in alcoholic patients24. Folate deficiencies are also commonly observed in

African HIV+ patient populations25. The results of the present study suggest that deficiencies in folate metabolism may play a role in the development of HIV-SN. This study also may be informative for other peripheral neuropathies, e.g. Type 2 diabetic neuropathy, and warrants additional clinical and experimental study.

A common NRTI toxicity is peripheral neuropathy, thought to be caused by mitochondrial toxicity26. NRTIs compete with endogenous nucleotides during mitochondrial DNA (mtDNA) replication, which leads to errors in mtDNA replication and depletion of mtDNA. This effect is seen more with older NRTIs, such as AZT and d4T, due to their higher affinity for the mitochondrial polymerase, polγ27. Several candidate gene studies have been performed to investigate the effect of host genetics on the development of NRTI-SN, however, few associations have been observed28-30. A

GWAS was performed in this dissertation to characterize unknown genetic predictors of

NRTI-SN. The SNP with the lowest p-value in this study was an intergenic SNP nearest to the VAMP4 gene, however, it did not reach genome-wide significance, likely due to a small study population. VAMP4 is largely uncharacterized, but has been implicated in the regulation of asynchronous synaptic transmission. This VAMP4 SNP is in an active regulatory region and is associated with a decrease in VAMP4 expression but replication was not successful in the small sample set available. A candidate gene study was also performed and SNPs in SLC22A1 and SLC28A1 were associated with

158 the development of NRTI-SN. These genes are known to transport NRTIs and may influence the exposure of a patient to NRTIs. The findings in the GWAS and candidate gene studies are interesting, although the sample size was limited, and suggest that further research into these genes is necessary.

The research in this dissertation highlights the importance of host genetics in HIV infection, ARV pharmacokinetics and toxicity. It also is important to study the effects of host genetics in multiple ethnic populations due to the differences in genetic variation observed in different ethnicities. By better understanding the role of host genetics, advances can be made in the prevention of disease complications and drug toxicities.

The results presented in this dissertation provide novel targets, but require additional experimental and clinical study.

159

6.1 References

1. UNAIDS, GLOBAL REPORT: UNAIDS report on the global AIDS epidemic 2013. (2013). 2. J. A. Moss, HIV/AIDS Review. Radiologic technology 84, 247-267; quiz p.268- 270 (2013). 3. S. G. Deeks, S. R. Lewin, D. V. Havlir, The end of AIDS: HIV infection as a chronic disease. Lancet 382, 1525-1533 (2013)10.1016/S0140-6736(13)61809- 7). 4. A. C. Justice, HIV and aging: time for a new paradigm. Current HIV/AIDS reports 7, 69-76 (2010)10.1007/s11904-010-0041-9). 5. J. Hemelaar, The origin and diversity of the HIV-1 pandemic. Trends in molecular medicine 18, 182-192 (2012)10.1016/j.molmed.2011.12.001). 6. FDA, Antiretroviral drugs used in the treatment of HIV infection. 7. W. H. Organization. (2013). 8. S. Broder, The development of antiretroviral therapy and its impact on the HIV- 1/AIDS pandemic. Antiviral research 85, 1-38 (2010)10.1016/j.antiviral.2009.10.002.The). 9. L. Dickinson, S. Khoo, D. Back, Pharmacokinetics and drug-drug interactions of antiretrovirals: an update. Antiviral research 85, 176-189 (2010)10.1016/j.antiviral.2009.07.017). 10. R. D. Moore, J. C. Keruly, CD4+ cell count 6 years after commencement of highly active antiretroviral therapy in persons with sustained virologic suppression. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 44, 441-446 (2007)10.1086/510746). 11. a. Telenti, U. M. Zanger, Pharmacogenetics of anti-HIV drugs. Annual review of pharmacology and toxicology 48, 227-256 (2008)10.1146/annurev.pharmtox.48.113006.094753). 12. V. Tozzi, Pharmacogenetics of antiretrovirals. Antiviral research 85, 190-200 (2010)10.1016/j.antiviral.2009.09.001). 13. A. Owen, M. Pirmohamed, S. H. Khoo, D. J. Back, Pharmacogenetics of HIV therapy. Pharmacogenetics and genomics 16, 693-703 (2006)10.1097/01.fpc.0000236338.41799.57). 14. S. Rodríguez-Nóvoa, P. Barreiro, I. Jiménez-Nácher, V. Soriano, Overview of the pharmacogenetics of HIV therapy. The pharmacogenomics journal 6, 234-245 10.1038/sj.tpj.6500374). 15. P. Riska, M. Lamson, T. Macgregor, J. Sabo, S. Hattox, J. Pav, J. Keirns, DISPOSITION AND BIOTRANSFORMATION OF THE ANTIRETROVIRAL DRUG NEVIRAPINE IN HUMANS ABSTRACT :. Pharmacology 27, (1999).

160

16. G. Ramachandran, K. Ramesh, A. K. Hemanth Kumar, I. Jagan, M. Vasantha, C. Padmapriyadarsini, G. Narendran, S. Rajasekaran, S. Swaminathan, Association of high T allele frequency of CYP2B6 G516T polymorphism among ethnic south Indian HIV-infected patients with elevated plasma efavirenz and nevirapine. The Journal of antimicrobial chemotherapy 63, 841-843 (2009)10.1093/jac/dkp033). 17. S. G. Heil, M. E. V. D. Ende, P. W. Schenk, I. V. D. Heiden, J. Lindemans, D. Burger, R. H. N. V. Schaik, Associations Between ABCB1, CYP2A6, CYP2B6, CYP2D6, and CYP3A5 Alleles in Relation to Efavirenz and Nevirapine Pharmacokinetics in HIV-Infected Individuals. Therapeutic drug monitoring 34, 153-159 (2012). 18. S. R. Penzak, G. Kabuye, P. Mugyenyi, F. Mbamanya, V. Natarajan, R. M. Alfaro, C. Kityo, E. Formentini, H. Masur, Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients in Uganda. HIV medicine 8, 86-91 (2007)10.1111/j.1468-1293.2007.00432.x). 19. T. Mahungu, C. Smith, F. Turner, D. Egan, M. Youle, M. Johnson, S. Khoo, D. Back, a. Owen, Cytochrome P450 2B6 516G-->T is associated with plasma concentrations of nevirapine at both 200 mg twice daily and 400 mg once daily in an ethnically diverse population. HIV medicine 10, 310-317 (2009)10.1111/j.1468-1293.2008.00689.x). 20. C. Ciccacci, P. Borgiani, S. Ceffa, E. Sirianni, M. C. Marazzi, A. M. D. Altan, G. Paturzo, P. Bramanti, G. Novelli, L. Palombi, Nevirapine-induced hepatotoxicity and pharmacogenetics: a retrospective study in a population from Mozambique. Pharmacogenomics 11, 23-31 (2010). 21. David W. W. Haas, T. Gebretsadik, G. Mayo, Usha N. N. Menon, Edward P. P. Acosta, A. Shintani, M. Floyd, C. M. M. Stein, Grant R. R. Wilkinson, Associations between CYP2B6 polymorphisms and pharmacokinetics after a single dose of nevirapine or efavirenz in African americans. The Journal of infectious diseases 199, 872-880 (2009)10.1086/597125). 22. D. E. Vance, M. Mugavero, J. Willig, J. L. Raper, M. S. Saag, Aging With HIV: A Cross-Sectional Study of Comorbidity Prevalence and Clinical Characteristics Across Decades of Life. The Journal of the Association of Nurses in AIDS Care : JANAC, 1-9 (2010)10.1016/j.jana.2010.04.002). 23. R. J. Ellis, D. Rosario, D. B. Clifford, J. C. McArthur, D. Simpson, T. Alexander, B. B. Gelman, F. Vaida, A. Collier, C. M. Marra, B. Ances, J. H. Atkinson, R. H. Dworkin, S. Morgello, I. Grant, Continued high prevalence and adverse clinical impact of human immunodeficiency virus-associated sensory neuropathy in the era of combination antiretroviral therapy: the CHARTER Study. Archives of neurology 67, 552-558 (2010)10.1001/archneurol.2010.76). 24. D. J. Lanska, Chapter 30: historical aspects of the major neurological vitamin deficiency disorders: the water-soluble B vitamins. Handb Clin Neurol 95, 445- 476 (2010)10.1016/S0072-9752(08)02130-1). 25. A. Alani, O. Vincent, A. Adewumi, A. Titilope, E. Onogu, A. Ralph, C. Hab, Plasma folate studies in HIV-positive patients at the Lagos university teaching 161

hospital, Nigeria. Indian journal of sexually transmitted diseases 31, 99-103 (2010)10.4103/0253-7184.74995). 26. C. M. Bailey, K. S. Anderson, in Biochimica et biophysica acta. (2010), vol. 1804, pp. 1213-1222. 27. P. L. Anderson, T. N. Kakuda, K. A. Lichtenstein, The cellular pharmacology of nucleoside- and nucleotide-analogue reverse-transcriptase inhibitors and its relationship to clinical toxicities. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 38, 743-753 (2004)10.1086/381678). 28. A. R. Kallianpur, T. Hulgan, J. a. Canter, M. D. Ritchie, J. L. Haines, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. W. Haas, Hemochromatosis (HFE) gene mutations and peripheral neuropathy during antiretroviral therapy. AIDS (London, England) 20, 1503-1513 (2006)10.1097/01.aids.0000237366.56864.3c). 29. J. a. Canter, D. W. Haas, a. R. Kallianpur, M. D. Ritchie, G. K. Robbins, R. W. Shafer, D. B. Clifford, D. G. Murdock, T. Hulgan, The mitochondrial pharmacogenomics of haplogroup T: MTND2*LHON4917G and antiretroviral therapy-associated peripheral neuropathy. The pharmacogenomics journal 8, 71- 77 (2008)10.1038/sj.tpj.6500470). 30. H. Yamanaka, H. Gatanaga, P. Kosalaraksa, S. Matsuoka-Aizawa, T. Takahashi, S. Kimura, S. Oka, Novel mutation of human DNA polymerase gamma associated with mitochondrial toxicity induced by anti-HIV treatment. The Journal of infectious diseases 195, 1419-1425 (2007)10.1086/513872).

162

163